Synthesis and optimization of a new family of type 3 17β-hydroxysteroid dehydrogenase inhibitors by parallel liquid-phase chemistry

Article abstract of DOI:10.1021/jm010286y

Type 3 17β-hydroxysteroid dehydrogenase (17β-HSD) transforms 4-androstene-3,17-dione (Δ⁴-dione) into the androgen testosterone. To produce potent inhibitors of this key steroidogenic enzyme, we performed parallel liquid-phase synthesis of 3β-su

Full text of DOI:10.1021/jm010286y

640
J . Med. Chem. 2002, 45, 640-653
Syn th esis a n d Op tim iza tion of a New F a m ily of Typ e 3 17â-Hyd r oxyster oid
Deh yd r ogen a se In h ibitor s by P a r a llel Liqu id -P h a se Ch em istr y
Rene´ Maltais, Van Luu-The, and Donald Poirier*
Medicinal Chemistry Division, Oncology and Molecular Endocrinology Research Center, Centre Hospitalier Universitaire de
Que´bec (CHUQ) and Universite´ Laval, 2705 Laurier Boulevard, Sainte-Foy, Que´bec G1V 4G2, Canada
Received J une 27, 2001
Type 3 17â-hydroxysteroid dehydrogenase (17â-HSD) transforms 4-androstene-3,17-dione (∆4-
dione) into the androgen testosterone. To produce potent inhibitors of this key steroidogenic
enzyme, we performed parallel liquid-phase synthesis of 3â-substituted androsterone (ADT)
libraries (A-D) in good yields and average high-performance liquid chromatography (HPLC)
purities of 92-94%. The first library (A) of 3â-amidomethyl-ADT derivatives (168 members),
including two levels of molecular diversity on the amide (R1 and R2), was synthesized with a
parallel liquid-phase method (method I) in less time than with the classic chemistry method.
The screening of library A revealed that relatively small hydrophobic chains at R1 (5-8 carbons)
and small hydrophobic substituents at R2 (1-4 carbons) provided the most potent inhibitors.
In accordance with these inhibition results, a second library (B) of 3â-amidomethyl-ADT
derivatives (56 members) was generated in a very short time using an improved method based
on scavenger resins and liquid-phase parallel chemistry. Library B produced more potent
inhibitors than library A and provided useful structure-activity relationships that directed
the design of a third library (C) of 49 members. Once again, very potent inhibitors were identified
from library C and 3â-[(N-adamantylmethyl-N-butanoyl)aminomethyl]-3R-hydroxy-5R-an-
drostan-17-one (C-7-3) was identified as the most potent inhibitor of the three libraries with
an inhibitory activity (IC50 ) 35 nM) 18-fold higher than that of the natural substrate of the
enzyme, ∆4-dione, (IC50 ) 650 nM) used itself as inhibitor. Finally, we designed a library (D)
of 3-carbamate-ADT derivatives (25 members) using the efficient parallel liquid-phase method
III, which allowed the synthesis of more rigid molecules with two levels of molecular diversity
(R1/R2 and R3) in the local area occupied by the adamantane group of C-7-3. Interestingly, one
of the most potent inhibitors of library D, the 3R-spiro-{3′-[3′′-N-morpholino-2′′-(3′′′-cyclopentyl-
propionyloxy)propyl]-2′-oxo-oxazolidin-5′-yl}-5R-androstan-17-one (D-5-4), showed an inhibitory
activity on type 3 17â-HSD similar to that of compound C-7-3, while exhibiting a nonandrogenic
profile.
In tr od u ction
androgen receptor (AR) (antiandrogens) or both.16,17 The
optimal androgen blockade therapy presently used
consists of the administration of an LHRH agonist
(medical castration) in combination with a pure anti-
androgen (flutamide).12,18,19 However, recent studies
have shown that the widely used chemical castration
with depot LHRH agonist fails, in nearly 20% of cases,
to achieve castrate levels of T in men (<20 ng/dL).20,21
Also, the relatively weak affinity of the antiandrogen
flutamide for the AR22 leaves the residual androgens
(from testis or adrenals), free to interact with the AR
and to potentially activate the growth of prostate cancer
cells, which is especially harmful in the case of androgen
hypersensitive tumors.23 Thus, to achieve an optimal
blockade of androgen formation, we are interested in
developing a potent inhibitor of type 3 17â-HSD, which
will eventually be used as an adjuvant to an LHRH
agonist and/or an antiandrogen for prostate cancer
treatment.
Type 3 17â-hydroxysteroid dehydrogenase (17â-HSD)
is a steroidogenic enzyme that catalyzes the reduction
of 4-androstene-3,17-one (∆4-dione) to testosterone (T)
using reduced nicotinamide adenine dinucleotide phos-
phate (NADPH) as the cofactor.1-3 This third member
of the 17â-HSD enzyme family4 is found almost exclu-
sively in the testis and contributes to the production of
approximately 60% of total active androgens in men.5
The inhibition in the testis of type 3 17â-HSD and the
inhibition in peripheral tissues of other key enzymes
such as type 5 17â-HSD,6,7 5R-reductases,8-10 and
steroid sulfatase,11 represent interesting strategies
toward a complete blockade of the biosynthesis of
androgens T and dihydrotestosterone (DHT) (Figure 1).
In the case of hormone sensitive diseases such as
prostate cancer, obtaining a complete androgen blockade
is crucial to counter the proliferation effect of androgens
on prostate cancer cells.12-15 Many different endocrine
treatments are now available to block either the tes-
ticular source of androgens (medical or surgical castra-
tion) or the effect of androgens T and DHT on the
Preliminary studies on the inhibition of type 3 17â-
HSD identified androsterone (ADT) as a good lead
compound (IC50 ) 330 nM) (Figure 2).24 The subsequent
synthesis of ADT derivatives bearing various 3â hydro-
phobic chains gave more potent inhibitors than ADT.25
In fact, with an IC50 value of 57 nM, 3â-phenylmethyl-
* Author to whom correspondence should be addressed. Tel: (418)-
654-2296. Fax: (418)654-2761. E-mail: donald.poirier@crchul.ulaval.ca.
10.1021/jm010286y CCC: $22.00 © 2002 American Chemical Society
Published on Web 12/27/2001

Type 3 17â-Hydroxysteroid Dehydrogenase Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 641
F igu r e 1. Role of type 3 17â-HSD in the transformation of the inactive androgen 4-androstene-3,17-dione (∆4-dione) into active
androgens T and DHT. Various strategies (i-iii) to block their biosynthesis involve: (i) Inhibitors of steroid sulfatase (peripheral
tissues); (ii) inhibitors of type 3 (testis) or type 5 17â-HSDs (peripheral tissues); and (iii) inhibitors of 5R-reductases (peripheral
tissues). DHEAS, dehydroepiandrosterone sulfate; DHEA, dehydroepiandrosterone.
ADT (1) was the most potent type 3 17â-HSD inhibitor
up to now. This molecule is six times more powerful
than ADT and 13 times more powerful than ∆4-dione
(used as inhibitor) to inhibit the enzyme reduction of
labeled ∆4-dione (0.1 µM).25 To optimize this inhibitor,
we were interested in developing strategies for the rapid
generation of a large number of diversified 3â-substi-
tuted-ADT derivatives. Our plan was essentially to use
a combinatorial chemistry approach to generate librar-
ies of steroid derivatives by either liquid-phase or solid-
phase synthesis.26,27 Recently, we have published a
methodology for the solid-phase parallel synthesis of 3â-
peptido-3R-hydroxy-5R-androstan-17-ones (2).28,29 In ad-
dition to these 3â-peptido-steroids, we were also inter-
ested in obtaining compounds possessing a tertiary
alkylamide chain at position 3â of ADT (compound 3).
Such compounds could be more biologically stable than
peptido-steroids 2, which are potentially sensitive to
hydrolases (aminopeptidases, carboxypeptidases, and
dipeptidases). While designing a combinatorial route to
synthesize these desired 3â-amidomethyl-ADT deriva-
tives of general structure 3, we chose a 3â-oxirane
functionality as precursor of molecular diversity. Fol-
lowing our preliminary results,30 we now report a full
account of the design, chemical synthesis, and biological
evaluation of efficiency as type 3 17â-HSD inhibitors of
3â-amidomethyl-ADT libraries generated by efficient
liquid-phase parallel synthesis.
F igu r e 2. Steroidal inhibitors of type 3 17â-HSD derived from
ADT and previously reported by us (compounds 1 and 2) or
proposed here (compound 3).
Meth od I. As a model, a library of 20 members was
first synthesized to evaluate the versatility and limits
of our proposed liquid-phase approach (method I, in
Scheme 2).30 The first level of diversity (R1) was
introduced by a regioselective opening of the oxirane
precursor 4 by a series of four primary amines at reflux
in the presence of lithium perchlorate.33 This reaction
was carried out in parallel using separate reactors
(sealed vials), one for each different primary amine used
(4 × R1). The resulting solution was diluted with ethyl
acetate and washed with water to remove the perchlo-
rate salt. The organic layer was then evaporated under
reduced pressure, and the crude compound was quickly
filtered in a silica gel syringe to remove the excess of
primary amines and unreacted oxirane precursor 4 from
the secondary amine 10. The second level of molecular
diversity (R2) was then obtained by a regioselective
reaction of the four secondary amines 10, previously
divided in 20 vials, with five acyl chlorides (R2COCl) in
the presence of pyridine to give the amides 11. It was
then necessary to perform silica gel filtration to obtain
the 20 library members in a high average purity of 93%
as determined by high-performance liquid chromatog-
raphy (HPLC). Using the same method, a larger library
of 168 members (library A) was generated in an average
HPLC purity of 92% (Table 1) for a sampling of 12
members (using a root square random sampling).34
However, after the synthesis of library A, we realized
that some aspects of method I were not optimal, and it
was found to be too laborious for an automated process.
In fact, the repetitive silica gel filtration needed after
Resu lts a n d Discu ssion
Ch em istr y. Our choice to use the liquid-phase me-
dium for the production of the desired libraries became
clear after our disappointing attempts to link the key
steroid precursors 4 and 5 on a solid support (Scheme
1). Indeed, the coupling of 3â-oxirane-androstan-17-one
(4)30 on polymer-bound glycerol, using the Leznoff acetal
linker,31,32 revealed the instability of the oxirane in these
coupling conditions. Alternatively, we grafted the acetal
linker on steroid 4 to generate 5 and tried coupling it
with a Merrifield resin to obtain 6. Once again, the
oxirane was found instable in this coupling step. As a
last attempt, we then generated oxirane 6 directly in
the solid phase starting from the epiandrosterone linked
to polymer-bound glycerol (resin 7) and added two more
steps to obtain the model compound 8. As a result, we
recovered at the end of the solid-phase sequence (oxida-
tion, epoxidation, aminolysis, and acylation) a crude
mixture (25% yield) containing the desired ADT deriva-
tive 8 but also impurities originating mainly from
incomplete reactions. Consequently, we finally turned
our attention to a liquid-phase combinatorial methodol-
ogy to provide the library of desired 3â-alkylamide-3R-
hydroxy-5R-androstan-17-ones.

642 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Sch em e 1. Solid-Phase Assaysa
Maltais et al.
a
Reagents and conditions. (a) Steroid 4 or epiandrosterone (0.75 mmol), TMOF (0.75 mmol), Sc(OTf)3, polymer-bound glycerol (0.25
mmol), toluene. (b) Steroid 5, NaH, merrifield resin (1.0 mmol/g), DMF. (c) TPAP, NMO, molecular sieves, CH2Cl2. (d) NaH, Me3SOI,
DMSO. (e) Octylamine, LiClO4, CH3CN, 60 °C, room temperature. (f) Hexanoyl chloride, pyridine. (g) 2.0 N HCl in dioxane (containing
0.03% of H2O), room temperature.
Sch em e 2. Parallel Liquid-Phase Synthesis (Methods I and II)a
a
Reagents and conditions. (a) NaH, Me3SOI, DMSO. (b) TPAP, NMO, molecular sieves, CH2Cl2. (c) Ethylene glycol, p-TSA, toluene,
reflux with Dean-Stark. Method I: (d) (i) R1NH2, LiClO4, CH3CN, 60 °C, room temperature; (ii) SiO2 filtration. (e) (i) R2COCl, pyridine;
(ii) H2O washing and then SiO2 filtration. Method II: (d) (i) R1NH2, EtOH, 60 °C; (ii) acetone/HCl 3 M (1:1); (iii) SiO2 filtration. (e) (i)
R2COCl, piperidinomethyl polystyrene (3.6 mmol/g); (ii) aminomethyl polystyrene (1.0 mmol/g), filtration.
the acylation step was identified as the most time-
consuming step especially when the number of com-
pounds generated is high. Thus, some modifications
were necessary to avoid this limiting step and to
improve this liquid-phase approach.
in ethanol rather than an amine and LiClO4 in aceto-
nitrile, thus limiting the workup procedure to a simple
evaporation of ethanol. The dioxolane protective group
was then easily removed with a 3 N HCl aqueous
solution in acetone, followed by a basic workup (10%
NaHCO3). A simple filtration on silica gel then afforded
the desired steroid 10 in high purity. Thereafter, the
second level of diversity (R2) was added using an excess
of the acyl chloride (2 equiv) in dichloromethane with
an acid scavenger resin (piperidinomethyl polystyrene)
and a nucleophilic resin (aminomethylated polysty-
rene),35 based on CMR/R methodology both to mop up
the proton released during acylation and to eliminate
the excess of acid chloride, respectively. A simple
filtration then removed the scavenger resins. The
filtrate was then evaporated on a Speedvac apparatus
giving the desired steroids of general structure 11. Two
Meth od II. An unknown product (10-15%, by thin-
layer chromatography (TLC) analysis), suspected to be
a ketimine resulting from the reaction between the C17-
ketone and the primary amine, was found to contami-
nate steroid 10 generated by method I. In fact, a silica
gel filtration was necessary to remove this impurity and
to obtain the library A members in good purity. To avoid
the formation of undesirable byproducts, the carbonyl
group of epiandrosterone was first protected as a
dioxolane, an acid sensitive protective group, and then
transformed into library precursor 9. Steroid 9 was
submitted to aminolysis using only an amine (R1NH2)

Type 3 17â-Hydroxysteroid Dehydrogenase Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 643
Ta ble 1. Characterization of Sampled Members of Libraries
A-D
into terminal epoxide 17 by the one pot procedure
reported by Kolb and Sharpless.37
average
HPLC
The first level of diversity (R1/R2) was introduced by
an aminolysis of the oxirane using a secondary amine
(R1R2NH) and lithium perchlorate as reported above.
The excess of secondary amine reagent was removed
from 18 using a scavenger (methylisocyanate polysty-
rene), and the perchlorate salt was eliminated by a
washing step with water. The diastereoisomeric ratio
of the resulting â-amino-alcohol 18 was determined by
the formation of a Mosher ester38 and, as expected, was
found to be 50:50. The choice to generate a racemic
mixture rather than a pure chiral compound was herein
advantageous considering that two compounds were
screened at once. If however necessary, it is possible to
perform a J acobsen kinetic resolution,39 in order to
obtain a pure chiral epoxide 17, as the precursor of
library member in a pure diastereoisomeric form. From
racemic secondary alcohol 18, the second level of
diversity (R3) was obtained by using the optimized
acylation conditions of method II. A 25-membered
library of type D (compounds 19) was thus generated
to validate method III (Table 1) and to provide the
targeted inhibitors.
theoret obsvda library HPLC
timec purity
library compd
mass
mass
size
purityb (days) (%)b
method I
A
A
A
A
A
A
A
A
A
A
A
A
A-1-8
A-2-11
A-3-9
A-4-1
A-5-3
A-6-2
A-7-14
A-8-6
A-9-13
543.5
471.4
599.5
487.4
529.5
529.5
571.5
613.5
627.6
544.4
472.6
600.7
488.4
530.8
530.8
572.7
614.3
628.8
726.5
656.7
654.4
97
85
84
90
97
96
94
99
97
91
90
91
168
92
25
A-10-10 725.7
A-11-5 655.6
A-12-12 653.6
method II
480.4
B
B
B
B
B
B
B
B
B
B
B
B
B
B-2-4
B-2-7
B-3-6
B-4-1
B-4-4
B-5-1
B-5-6
B-6-1
B-6-6
B-6-7
B-7-1
B-8-2
B-8-4
479.2
429.3
455.3
419.3
497.2
447.3
473.4
475.4
501.4
503.4
441.3
469.2
479.2
94
97
82
97
95
97
94
90
82
93
99
92
91
430.6
456.6
420.3
498.3
448.5
474.6
476.3
502.4
504.6
442.6
470.6
480.4
56
94
4
Scr een in g of Libr a r ies A-C for In h ibitor y Ac-
tivity on Typ e 3 17â-HSD. Libr a r y A. Library A was
designed to explore the hydrophobic region of the
enzyme, a region previously identified by inhibition with
3â-alkyl-ADT derivatives. A library of 168 members (12
× 14), containing various alkylamines and aliphatic acyl
chlorides building blocks, was generated by method I
with the aim of finding the optimal carbon chain length
at R1 and R2. A selection of 40 representative members
of library A was chosen for a preliminary inhibition
screening (Figure 3). The evaluation of the ability of
selected members to inhibit the type 3 17â-HSD activity
transfected in human embryonic kidney (HEK)-293 cells
(homogenized) was done by measuring the amount of
labeled T formed from the labeled natural substrate ∆4-
dione in the presence of NADPH as the cofactor. The
results were expressed as inhibitory activity (percent)
at a concentration of 3 µM. An obvious observation was
that the long alkyl chains (nine carbons or more at R1
and five carbons or more at R2) were generally not well-
tolerated by the enzyme. In fact, the compounds pos-
sessing a chain length smaller than eight carbons at R1
and smaller than four carbons at R2 (compounds A-1-1,
A-2-2, A-3-3, and A-4-4) showed the strongest inhibi-
tion. Also, a short ramified chain at R2 (tert-butyl,
ethylhexyl, cyclobutyl, and cyclopropyl) seems to be well-
tolerated by the enzyme (A-1-14, A-2-13, A-3-12, and
A-4-11). Interestingly, the combination of a long chain
at R1 and of a short one at R2 (compound A-1-7) was
also found to give a good inhibition. Another important
observation was that the compounds with a single level
of diversity, without any R2 chains (free secondary
amine of general structure 10), showed small inhibition
values, bringing us to prefer the tertiary amide over the
secondary amine (results not reported). Thus, this first
library A provided useful structure-activity relation-
ships (SAR) information about the dimension of the
hypothetical hydrophobic pocket of the enzyme and
method II
C
C
C
C
C
C
C
C
C-2-6
C-3-7
C-4-3
C-4-5
C-5-1
C-6-1
C-6-2
C-7-7
491.3
507.4
615.3
671.4
465.3
511.3
525.4
537.4
492.5
508.1
616.4
672.6
466.4
512.6
526.6
538.7
86
99
86
97
98
98
96
76
49
92
4
4
method III
543.5
D
D
D
D
D
D-1-1
D-2-2
D-3-3
D-4-4
D-5-5
542.4
612.5
576.4
610.4
648.4
>80d
>80d
>80d
>80d
>80d
613.2
577.4
611.1
649.5
25
>80d
a
Low-resolution mass spectra (LRMS) in positive mode (M +
H)+. HPLC purity of steroid derivatives released from resin
b
(without purification steps). c Time required by one worker for the
d
synthesis of the library. Purity estimated to be higher than 80%
by TLC and 1H NMR analysis.
libraries (B and C) of 56 and 49 members were obtained
with this method in high average purity (94 and 92%,
respectively). Regarding the time required for preparing
different libraries (Table 1), method II was found more
advantageous than method I.
Meth od III. The parallel synthesis of the library D
members of general structure 19 (Scheme 3) required
a different methodology than that developed to generate
libraries A-C (methods I and II). The synthesis of
derivatives 19 first necessitates the synthesis of steroid
precursor 17. The previously described 3â-oxirane-5R-
androstan-17-one (4)30 was submitted to the aminolysis
conditions of method II using allylamine in ethanol at
60 °C to generate olefin 12. The cyclic carbamate of
compound 14 was then obtained in treating the amino
alcohol 12 with trisphosgene reagent36 in the presence
of diisopropylethylamine (DIPEA). Thereafter, the olefin
of compound 14 was oxidized using N-methyl morpho-
line-N-oxide (NMO) with a catalytic amount of osmium
tetroxide to give diol 16, which was directly converted

644 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Maltais et al.
Sch em e 3a
a
Reagents and conditions. (a) Allylamine, EtOH, 60 °C. (b) Triphosgene, DIPEA, CH2Cl2. (c) OsO4, NMO, t-BuOH, THF, H2O, (7:3:1).
(d) (i) MeC(OMe)3, PPTS, CH2Cl2; (ii) AcBr, Et3N, CH2Cl2; (iii) K2CO3, MeOH. Method III: (e) R1R2NH, LiClO4, CH3CN, 60 °C. (f)
Methylisocyanate polystyrene (1.0 mmol/g), CH2Cl2, filtration, H2O washing. (g) R3COCl, piperidinomethyl polystyrene (3.6 mmol/g),
dimethylaminopyridine polystyrene (1.0 mmol/g), EtOAc. (h) Aminomethyl polystyrene (1.0 mmol/g), CH2Cl2, filtration. (i) NaHCO3 (10%),
EtOAc, silica gel filtration.
F igu r e 3. Inhibition of type 3 17â-HSD (percent) by a selection of 40 members of library A at a concentration of 3 µM.
provided us with useful guidelines for elaborating new
libraries with novel building blocks on the tertiary
amide.
Libr a r y B. On the basis of what we learned from
library A, library B was synthesized with various types
of building blocks with a maximum length of eight
atoms at R1 and of four atoms at R2. Also, the library
was designed as to determine the relative importance
of certain parameters such as the polarity of the R2
chain (halogen vs methylene) and/or the rigidity and
density of the carbon chain at R1 (cyclohexyl, cyclopen-
tyl, cyclopropyl, and ether). The percentage of inhibition
was measured at a concentration of 0.3 µM for all library
members (Figure 4), yielding a wide range of inhibition
values ranging from 17 to 93%. Important SAR infor-
mation was acquired from this screening. It was obvious
that a cyclohexylmethyl group at R1 produced the
highest inhibition for most R2 groups (B-1-7, B-1-6, B-1-

Type 3 17â-Hydroxysteroid Dehydrogenase Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 645
F igu r e 4. Inhibition of type 3 17â-HSD by 56 members of library B at a concentration of 0.3 µM.
F igu r e 5. Inhibition of type 3 17â-HSD by 49 members of library C at a concentration of 0.3 µM.
5, B-1-3, and B-1-2). These compounds were also better
inhibitors than the best library A member (A-4-11)
when compared under the same conditions. Such results
suggest that a more rigid carbon chain at R1 increases
the affinity for the enzyme. On the other hand, a more
polar carbon chain bearing an ether function was not
advantageous as compared to the hydrophobic chains.
With these results in hand, we were interested to
generate a new library exploring these benefic param-
eters (library C).
Libr a r y C. On the basis of the results obtained from
library B, different, more rigid building blocks, such as
adamantylmethyl, were used at R1 to mimic the impor-
tant hydrophobic effect of a cyclohexylmethyl group
(Figure 5). However, small hydrophobic chains were
used at R2 but different ones than in libraries A and B.

646 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Maltais et al.
Ta ble 2. IC50 Values of Selected Inhibitors of Libraries A-D and Their Proliferative Activity on Shionogi (AR+) Cells
a
Proliferative activity (%) induced by 1 and 0.1 µM of tested compound on Shionogi (AR+) cells. Results are compared to the stimulation
b
induced by 0.3 nM of androgen DHT (100%). N ) not determined. Synthesis of compound 20 was realized from compound 9 in a 4 step
sequence of reactions (aminolysis, acylation, reduction of amide, and deprotection of acetal).
The results showed that the adamantylmethyl group
clearly provides the highest inhibition at R1, especially
in the presence of small chains at R2 (C-7-2, C-7-3, C-7-
4, and C-7-7). Also, the compound C-4-4 combining a
3,5-(ditrifluoromethyl)benzyl group at R1 and a trichloro
acetyl group at R2 gave strong inhibition. All of these
compounds were better inhibitors than the best library
B member (B-1-6).
we were curious about the effect of a rigidification of
the 3â-amidomethyl-3R-hydroxy system. We thus elabo-
rated a new strategy to quickly and easily generate a
library (D) of cyclic carbamate derivatives of general
structure 19 (Scheme 3). In addition to their rigid
structure, these cyclic carbamate derivatives had the
advantage of generating a local diversity in the same
area that the adamantane group occupied in the library
C. Such local diversity could generate additional ben-
efical interactions in this enzyme area. From the inhibi-
tion results of library D (Figure 6), it can be seen that
the morpholine (D-5-2, D-5-4, and D-5-5) and dipropyl-
amine building blocks (D-2-3, D-2-4, and D-2-5) at R1
and tert-butylphenyl at R2 (D-1-5, D-2-5, D-3-5, D-4-5,
and D-5-5) provide the strongest inhibition at 0.3 µM
(80-90%), with values similar to the inhibition obtained
with the best inhibitors of libraries A-C (A-4-11, B-1-
6, and C-7-3). However, the rigidification of the 3R-OH
and 3â-tertiary amide into a five-membered cyclic
carbamate seems to slightly decrease the inhibition, as
illustrated by a comparison between the compound
C-7-3 (noncyclic version) and the compound 15 (car-
bamate cyclic version). The IC50 of one of the best
compounds of library D (D-5-4) was determined in the
same assay as the best inhibitor of libraries A-C (C-
7-3). The IC50 values of the two compounds (57 and 74
nM) are very close, and both are better than the IC50 of
reference compound 1 (98 nM) (Table 2). Considering
that the compounds of library D are mixtures of two
diastereoisomers, the pure diastereoisomeric form of
active compounds, like D-5-4, could possibly exert an
even stronger inhibitory activity.
Deter m in a tion of IC50 Va lu es of th e Best In h ibi-
tor s Iden tified fr om Libr ar ies A-C. After the screen-
ing of libraries A-C, we were interested in determining
the IC50 values of the best inhibitors of each library and
to compare them with other known inhibitors (Table 2).
One compound was chosen per library (A-4-11, B-1-6,
and C-7-3), and their IC50 values were determined in
the same enzymatic assay. The results showed that
compound C-7-3 bearing an adamantylmethyl group at
R1 was the most potent inhibitor of the three libraries
(IC50 ) 35 nM). Its IC50 value is approximately 18 times
higher than the IC50 value of the natural substrate ∆4-
dione (IC50 ) 650 nM) used itself as an inhibitor.
Complementary to these results, the reduction of the
amide function to an amine was done to verify the
influence of the tertiary amide vs the tertiary amine on
inhibition. The tertiary amine 20 (IC50 ) 80 nM) was
found to be only half as potent as the corresponding
amide C-7-3. This suggests a positive effect of the
carbonyl of the amide on inhibition. We therefore
suspect the formation of a hydrogen bond between the
carbonyl of amide derivatives and the 3R-alcohol that
could favoring the affinity of the compound C-7-3 for
the enzyme.
Scr een in g of Libr a r y D for In h ibitor y Activity
on Typ e 3 17â-HSD. Given the biological results of
libraries A-C and the hypothesis of a hydrogen bond,
An d r ogen ic Activity of P oten t In h ibitor s. A type
3 17â-HSD inhibitor to be used in prostate cancer
treatment obviously must be nonandrogenic. Consider-

Type 3 17â-Hydroxysteroid Dehydrogenase Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 647
F igu r e 6. Inhibition of type 3 17â-HSD by 25 members of library D at a concentration of 0.3 µM.
ing their C19-steroid androgenic nucleus, we thus tested
the two best inhibitors from our study (C-7-3 and D-5-
4) to evaluate their possible androgenic activity using
the model of androgen sensitive (AR+) Shionogi cells at
two concentrations (1 and 0.1 µM). As shown in Table
2, compound C-7-3 induced proliferation (96%) at 1 µM
that was of the same order than compound 1 (100%),
the best first generation inhibitor. On the other hand,
the compound D-5-4 displayed only a slight, not sig-
nificant proliferative activity at 1 and 0.1 µM. Using this
second criteria, inhibitor D-5-4 was thus the best
candidate from all generated libraries (A-D) showing
both a good inhibitory effect on type 3 17â-HSD and a
nonandrogenic profile.
that of natural substrate ∆4-dione (IC50 ) 650 nM) used
itself as an inhibitor. We finally designed a library (D)
of 3-carbamate-ADT derivatives (25 members) using the
efficient parallel liquid-phase method III, which allowed
the preparation of more rigid molecules with two levels
of molecular diversity (R1/R2 and R3) in the local area
occupied by the adamantane group in library C. One of
the most potent inhibitors of library D, the compound
D-5-4, showed an inhibitory activity similar to that of
compound C-7-3, combined to a nonandrogenic profile.
Exp er im en ta l Section
Gen er a l In for m a tion . DHT and epiandrosterone were
purchased from Steraloids (Wilton, NH). Scavenger resins
(methylisocyanate polystyrene, aminomethylated polystyrene,
and piperidinomethyl polystyrene) were purchased from Cal-
biochem-Novabiochem Corp. (San Diego, CA), and the polymer-
bound reagent ((dimethylamino)pyridine polystyrene, DMAP-
PS) was purchased from Argonaut Technologies (SanCarlos,
CA). Chemical reagents were purchased from Sigma-Aldrich
Canada Ltd. (Oakville, ON, Canada). The usual solvents were
obtained from Fisher Scientific (Montre´al, QC, Canada) and
were used as received. Anhydrous dichloromethane (CH2Cl2),
acetonitrile (CH3CN), dimethyl sulfoxide (DMSO), toluene, and
pyridine were obtained from Sigma-Aldrich. Anhydrous tet-
rahydrofuran (THF) was distilled from sodium/benzophenone
ketyl and kept under argon. Libraries A-C were synthesized
using sealed vials (10 mL), and library D was synthesized with
an ACT LabTech manual synthesizer (Advanced ChemTech,
Louisville, KY) using a 40 wells liquid-phase reaction block.
Silica gel (SiO2) filtration was performed using either 10 or
20 mL disposable polypropylene syringe (Sigma-Aldrich) using
a cotton plug and sand at the end of the syringe. TLC and
flash-column chromatography were performed on 0.20 mm
silica gel 60 F254 plates with 230-400 mesh ASTM silica gel
60, respectively (E. Merck, Darmstadt, Germany). The purity
of a random sampling of library was determined by HPLC
(Waters Associates, Milford, MA) using a NovaPak C18
reversed phase column (150 mm × 3.9 mm id) and an
ultraviolet detector (205 nM). Infrared spectra were recorded
on a Perkin-Elmer series 1600 Fourier transform infrared
spectrometer (Norwalk, CT), and only the significant bands
were reported in cm-1. Nuclear magnetic resonance spectra
Con clu sion
To produce potent inhibitors of type 3 17â-HSD, we
performed parallel liquid-phase synthesis of 3â-substi-
tuted-ADT libraries (A-D) in good yields and an aver-
age HPLC purity of 93%. The first library (A) of
3â-amidomethyl-ADT derivatives (3) (168 members),
including two levels of molecular diversity on the amide
(R1 and R2), was synthesized by a parallel liquid-phase
method I faster (25 days) than with classic chemistry.
The screening of library A revealed that relatively small
hydrophobic chains at R1 (5-8 carbons) and small
hydrophobic substituents at R2 (1-4 carbons) were two
characteristics of the most potent inhibitors. According
to these inhibition results, a second library (B) of 3â-
amidomethyl-ADT derivatives (56 members) was gener-
ated in a very short time (4 days) using the improved
method II based on scavenger resins and liquid-phase
parallel chemistry. Library B generated more potent
inhibitors than library A and provided useful informa-
tion on SAR that was used for the design of a new
library (C). Once again, library C (49 members) showed
very potent inhibitors; compound C-7-3 was identified
as the most potent inhibitor of libraries A-C, with an
inhibitory activity (IC50 ) 35 nM) 18-fold higher than

648 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Maltais et al.
(NMR) were recorded at 300 MHz for 1H and 75.5 MHz for
13C on a Brucker AC/F300 spectrometer (Billerica, MA) and
reported in parts per million. Low-resolution mass spectra
(LRMS) were recorded on a PE Sciex API-150ex apparatus
(Foster City, CA) equipped with a turbo ionspray source.
CH2N-3â), 4.60 (s, OH). LRMS for C31H54NO3 [M + H]+: 488.4
m/z. HPLC (CH3CN:H2O:CH3OH, 35:20:45): 90% of purity.
A-5-3. 3â-[(N-Nonyl-N-pentanoyl)aminomethyl]-3R-hydroxy-
5R-androstan-17-one. IR (KBr): 3416 (OH), 1736 (CdO,
ketone), 1611 (CdO, amide). 1H NMR (CDCl3): 0.78 (s, 3H,
CH3-19), 0.85 (s, 3H, CH3-18), 0.88 (t, J ) 6.3 Hz, 3H, CH3-
CH2), 0.92 (t, J ) 7.3 Hz, 3H, CH3CH2), 1.00-2.20 (39H), 2.34
(t, J ) 7.60 Hz, 2H, CH2CO), 2.42 (dd, J 1 ) 8.6 Hz, J 2 ) 19.1
Hz, 1H of CH2-16), 3.30 (m, 4H, CH2CH2N and CH2N-3â), 4.65
(s, OH). LRMS for C34H60NO3 [M + H]+: 530.8 m/z. HPLC
(CH3CN:H2O:CH3OH, 35:10:55): 97% of purity.
A-6-2. 3â-[(N-Decyl-N-butanoyl)aminomethyl]-3R-hydroxy-
5R-androstan-17-one. IR (film): 3306 (OH), 1739 (CdO, ke-
tone), 1622 (CdO, amide). 1H NMR (CDCl3): 0.78 (s, 3H, CH3-
19), 0.85 (s, 3H, CH3-18), 0.88 (t, J ) 6.4 Hz, 3H, CH3CH2),
0.96 (t, J ) 7.3 Hz, 3H, CH3(CH2)2CO), 1.00-2.10 (39H), 2.32
(t, J ) 7.6 Hz, 2H, CH2CO), 2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.1
Hz, 1H of CH2-16), 3.31 (m, 4H, CH2N-3â), 4.63 (s, OH). LRMS
for C34H60NO3 [M + H]+: 530.8 m/z. HPLC (CH3CN:H2O:CH3-
OH, 35:10:55): 96% of purity.
P r ep a r a tion of Libr a r y A Mem ber s (Meth od I). 3â-
Am in om eth yl-3r-h yd r oxy-5r-a n d r osta n -17-on es (10). In
12 vials (15 mL) were added in parallel the spiro-3(R)-oxirane-
5R-androstan-17-one (4)30,40 (363 mg, 1.2 mmol), dry CH3CN
(10 mL), anhydrous lithium perchlorate (255 mg, 2.4 mmol),
and the primary amines (1.2 mmol) individually. The mixtures
were heated in sealed vials at 60 °C for 7 days under an
atmosphere of argon, and the solutions were evaporated under
vacuum. The crude products were dissolved in ethyl acetate
(30 mL), washed with water (150 mL), and then purified by a
fast filtration on silica gel (10 mL) in polypropylene syringes
of 20 mL using EtOAc: hexanes (1:1) as the eluent. After the
solvent was evaporated under reduced pressure, secondary
amines 10 (see Figure 3 for R1) were obtained in acceptable
yields (60-80%).
A-7-14. 3â-[(N-Undecanyl-N-tert-butylcarbonyl)aminomethyl]-
3R-hydroxy-5R-androstan-17-one. IR (KBr): 3432 (OH), 1743
(CdO, ketone), 1609 (CdO, amide). 1H NMR (CDCl3): 0.78
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-19), 0.88 (t, J ) 6.3 Hz, 3H,
CH3CH2), 1.07 (s, 9H, (CH3)3-C), 1.00-2.20 (40H), 2.24 (s, 2H,
CH2CO), 2.42 (dd, J 1 ) 8.6 Hz, J 2 ) 19.2 Hz, 1H of CH2-16),
3.32 (t, J ) 7.7 Hz, 4H, CH2CH2N and CH2N-3â), 4.76 (s, OH).
LRMS for C37H66NO3 [M + H]+: 572.70 m/z. HPLC (CH3CN:
H2O:CH3OH, 35:10:55): 94% of purity.
3â-Am idom eth yl-3r-h ydr oxy-5r-an dr ostan -17-on es (11).
Each of the secondary amines 10 obtained above was equally
divided in 14 different portions for the acylation reaction. The
appropriate acyl chlorides (0.05 mmol) were added individually
to the secondary amine (0.05 mmol) in the presence of
anhydrous pyridine (0.1 mmol) and CH2Cl2 (1 mL). After 3 h
at room temperature, the CH2Cl2 was removed under vacuum
(Speedvac apparatus, Winchester, VA) and ethyl acetate was
added. The organic layer was washed with an aqueous NaOH
solution (40%) and evaporated. The crude amides 11 were then
purified by a fast silica gel filtration (EtOAc:hexanes, 3:7)
using 10 g of silica gel in a 20 mL polypropylene syringe. The
first 10 mL eluted was discarded, and the next 20 mL
recovered was evaporated to give the 3â-amidomethyl-3R-
hydroxy-5R-androstan-17-ones (168 members) in acceptable
yield (40-70%) and a high average purity (92%) for a random
sampling of 12 members: A-1-8, A-2-11, A-3-9, A-4-1, A-5-3,
A-6-2, A-7-14, A-8-6, A-9-13, A-10-10, A-11-5, and A-12-12
(Table 1). See Figure 3 for the chemical structure of com-
pounds.
A-8-6. 3â-[(N-Dodecyl-N-octanoyl)aminomethyl]-3R-hydroxy-
5R-androstan-17-one. IR (KBr): 3422 (OH), 1739 (CdO,
ketone), 1614 (CdO, amide). 1H NMR (CDCl3): 0.78 (s, 3H,
CH3-19), 0.85 (s, 3H, CH3-19), 0.87 (t, J ) 8.0 Hz, 6H, 2 ×
CH3CH2), 1.00-2.47 (55H), 2.33 (t, J ) 6.7 Hz, 2H, CH2CO),
3.31 (m, 4H, CH2CH2N and CH2N-3â), 4.76 (s, OH). LRMS for
C
40H72NO3 [M + H]+: 614.3 m/z. HPLC (CH3CN:H2O:CH3OH,
25:30:45): 99% of purity.
A-9-13. 3â-[(N-Tridecyl-N-2′-ethylhexanoyl)aminomethyl]-
3R-hydroxy-5R-androstan-17-one. IR (film): 3386 (OH), 1741
(CdO, ketone), 1617 (CdO, amide). 1H NMR (CDCl3): 0.78
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.88 (t, J ) 6.7 Hz, 6H,
2 × CH3CH2), 0.90 (t, J ) 6.8 Hz, 3H, CH3CH2), 1.00-2.20
(50H), 2.44 (m, 2H, CH2CO and 1H of CH2-16), 3.33 (m, 4H,
CH2CH2N and CH2N-3â), 4.8 (s, OH). LRMS for C41H74NO3
[M + H]+: 628.8 m/z. HPLC (CH3CN:H2O:CH3OH, 55:10:35):
97% of purity.
A-1-8. 3â-[(N-Pentyl-N-decanoyl)aminomethyl]-3R-hydroxy-
5R-androstan-17-one. IR (film): 3374 (OH), 1741 (CdO, ke-
tone), 1620 (CdO, amide). 1H NMR (CDCl3): 0.78 (s, CH3-19),
0.85 (s, 3H, CH3-18), 0.89 and 0.92 (2t, 6H, J ) 6.9 Hz, 2 ×
CH3CH2), 2.33 (t, 2H, J ) 7.5 Hz, CH2CO), 1.00-2.20 (42H),
2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.1 Hz, 1H of CH2-16), 3.31 (m,
A-10-10. 3â-[(N-Tetradecyl-N-tetradecanoyl)aminomethyl]-
3R-hydroxy-5R-androstan-17-one. IR (film): 3371 (OH), 1742
(CdO, ketone), 1620 (CdO, amide). 1H NMR (CDCl3): 0.78
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.88 (t, J ) 7.2 Hz, 6H,
2 × CH3CH2), 1.00-2.20 (67H), 2.33 (t, J ) 7.6 Hz, 2H, CH2-
CO), 2.42 (dd, J 1 ) 8.5 Hz, J 2 ) 19.1 Hz, 1H of CH2-16), 3.32
(m, 4H, CH2CH2N and CH2N-3â), 4.76 (s, OH). LRMS for
4H, CH2CH2N and CH2N-3â), 4.64 (s, OH). LRMS for C35H62
-
NO3 [M + H]+: 544.4 m/z. HPLC (CH3CN:H2O:CH3OH, 35:
10:55): 97% of purity.
A-2-11. 3â-[(N-Hexyl-N-cyclopropylcarbonyl)aminomethyl]-
3R-hydroxy-5R-androstan-17-one. IR (film): 3374 (OH), 1740
(CdO, ketone), 1610 (CdO, amide). 1H NMR (CDCl3): 0.78
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.88 (t, J ) 7.5 Hz, 3H,
CH3CH2), 1.00-2.10 (34H), 2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.2
Hz, 1H of CH2-16), 3.35 (qapp of AB system, 2H, CH2N-3â), 3.48
(m, 2H, CH2CH2N) 4.64 (s, OH). LRMS for C30H50NO3 [M +
H]+: 472.6 m/z. HPLC (CH3CN:H2O:CH3OH, 35:20:45): 85%
of purity.
C
48H88NO3 [M + H]+: 726.5 m/z. HPLC (CH3CN:H2O:CH3OH,
70:05:25): 91% of purity.
A-11-5. 3â-[(N-Hexadecyl-N-heptanoyl)aminomethyl]-3R-hy-
droxy-5R-androstan-17-one. IR (film): 3331 (OH), 1742 (Cd
O, ketone), 1621 (CdO, amide). 1H NMR (CDCl3): 0.78 (s, 3H,
CH3-19), 0.85 (s, 3H, CH3-18), 0.88 (t, 6H, J ) 5.7 Hz, 2 ×
CH3CH2), 1.00-2.20 (57H), 2.32 (t, J ) 7.6 Hz, 2H, CH2CO),
2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.1 Hz, 1H of CH2-16), 3.30 (m,
A-3-9. 3â-[(N-Heptyl-N-dodecanoyl)aminomethyl]-3R-hydroxy-
5R-androstan-17-one. IR (film): 3306 (OH), 1742 (CdO, ke-
tone), 1621 (CdO, amide). 1H NMR (CDCl3): 0.78 (s, 3H, CH3-
19), 0.85 (s, 3H, CH3-18), 0.87 and 0.89 (2t, 6H, J ) 6.6 Hz, 2
× CH3CH2), 1.00-2.20 (52H), 2.33 (t, J ) 7.6 Hz, 2H, CH2-
CO), 2.42 (dd, J 1 ) 8.6 Hz, J 2 ) 19.1 Hz, 1H of CH2-16), 3.31
(m, 4H, CH2CH2N and CH2N-3â), 4.65 (s, OH). LRMS for
4H, CH2CH2N and CH2N-3â), 4.64 (s, OH). LRMS for C43H78
-
NO3 [M + H]+: 656.7 m/z. HPLC (CH3CN:H2O:CH3OH, 55:
10:35): 90% of purity.
A-12-12. 3â-[(N-Octadecyl-N-cyclobutylcarbonyl)aminome-
thyl]-3R-hydroxy-5R-androstan-17-one. IR (film): 3374 (OH),
1735 (CdO, ketone), 1618 (CdO, amide). 1H NMR (CDCl3):
0.78 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.88 (t, J ) 6.3 Hz,
3H, CH3CH2), 1.00-2.50 (61H), 3.20 and 3.30 (2m, 4H,
CH2CH2N and CH2N-3â) 4.76 (s, OH). LRMS for C43H76NO3
[M + H]+: 654.4 m/z. HPLC (CH3CN:H2O:CH3OH, 55:10:35):
91% of purity.
C
39H70NO3 [M + H]+: 600.7 m/z. HPLC (CH3CN:H2O:CH3OH,
55:10:35): 84% of purity.
A-4-1. 3â-[(N-Octyl-N-propionyl)aminomethyl]-3R-hydroxy-
5R-androstan-17-one. IR (film): 3375 (OH), 1741 (CdO, ke-
tone), 1620 (amide, CdO). 1H NMR (CDCl3): 0.78 (s, 3H, CH3-
19), 0.85, (s, 3H, CH3-18), 0.86 (t, J ) 7.5 Hz, 3H, CH3CH2),
1.15 (t, J ) 7.3 Hz, 3H, CH3CH2), 1.00-2.20 (33H), 2.37 (m,
3H, CH2CO and 1H of CH2-16), 3.30 (m, 4H, CH2CH2N and
P r ep a r a tion of Libr a r ies B a n d C Mem ber s (Meth od
II). Syn th esis of Sp ir o-3(R)-oxir a n e-5r-a n d r osta n e-17-

Type 3 17â-Hydroxysteroid Dehydrogenase Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 649
d ioxola n e (9). A sequence of three steps was needed to obtain
compound 9 from epiandrosterone.
B-2-4. 3â-[(N-Cyclopropyl-N-bromoacetyl)aminomethyl]-3R-
hydroxy-5R-androstan-17-one. IR (film): 3432 (OH), 1738
(CdO, ketone), 1645 (CdO, amide). 1H NMR (CDCl3): 0.79
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.80-2.20 (25H), 2.43
(dd, J 1 ) 8.8 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.00 (m, 1H,
CHN), 3.16 (s, OH), 3.44 (s, 2H, CH2N-3â), 4.18 (s, 2H, CH2-
Br). LRMS for C25H39BrNO3 [M + H]+: 480.4 m/z (79Br) and
482.4 m/z (81Br). HPLC (CH3CN:H2O:CH3OH, 30:40:30): 94%
of purity.
(1) P r otection of C17-Keton e a s Dioxola n e. To a solution
of epiandrosterone (10.0 g, 34.4 mmol) in anhydrous toluene
(500 mL) were added ethylene glycol (19.2 mL, 344 mmol) and
p-toluenesulfonic acid (p-TSA) (654 mg, 3.4 mmol). The result-
ing solution was heated at reflux for 24 h under an atmosphere
of argon using a Dean-Stark apparatus. The solution was
poured in cold water (2 L) and extracted with EtOAc (2 × 500
mL) and washed with water (2 L). The organic layer was dried
over MgSO4 and evaporated under reduced pressure to give
12.0 g of crude 3â-hydroxy-5R-androstane-17-dioxolane.
(2) Oxid a tion of 3â-OH to Keton e. To the crude dioxolane
obtained above (12.0 g) in anhydrous dichloromethane (150
mL) were added molecular sieves (20 g), NMO (6.92 g, 39.4
mmol), and tetrapropylammonium perruthenate (TPAP) (692
mg, 1.96 mmol). The mixture was stirred for 2 h under an
atmosphere of argon and then directly filtered on a silica gel
plug (SiO2, 100 g) to give 10.5 g of 3-oxo-5R-androstane-17-
dioxolane.
(3) F or m a tion of 3-Sp ir o-(R)-oxir a n e. To a solution of
trimethylsulfoxonium iodide (13.9 g, 63.3 mmol) in dry DMSO
(350 mL) was carefully added sodium hydride 60% in oil (2.5
g, 62.5 mmol) under an atmosphere of argon. The solution was
stirred at room temperature for 1 h before adding the crude
3-oxo-5R-androstane-17-dioxolane (10.5 g, 31.6 mmol) dissolved
in dry THF (200 mL). The mixture was stirred overnight,
poured in ice/water (2 L), and extracted with EtOAc (2 × 300
mL). The combined organic layer was washed with brine, dried
over MgSO4, filtered, and evaporated to dryness. Purification
of the resulting crude product (10.0 g) by flash chromatography
(EtOAc:hexanes:triethylamine (TEA), 5:95:1) yielded 6.5 g
(54% for 3 steps) of compound 9. 1H NMR (CDCl3): 0.83 (s,
6H, CH3-19 and CH3-18), 2.59 (s, 2H, CH2O-3â), 0.80-2.10
(22H), 3.88 (m, 4H, OCH2CH2O). 13C NMR (CDCl3): 11.2, 14.3,
20.3, 22.5, 28.3, 29.1, 30.6, 31.0, 34.1, 35.4, 35.6, 35.7, 35.9,
43.6, 45.8, 50.2, 53.4, 53.6, 58.4, 64.5 (CH2O), 65.1 (CH2O),
119.4 (C-17). LRMS for C22H35O3 [M + H]+: 347.4 m/z.
3â-Am in om eth yl-3r-h ydr oxy-5r-an dr ostan -17-on es (10).
To the oxirane 9 (150 mg, 0.43 mmol) in ethanol (3 mL) were
added in parallel the appropriate primary amines (1.3 mmol),
and the mixtures were heated at 60 °C for 36 h. The resulting
solutions were evaporated, and the crude products obtained
were dissolved with acetone (5 mL). An aqueous HCl 3 N
solution (5 mL) was then added to each reactor, and solutions
were stirred for 2 h at room temperature. The acid mixtures
were poured in a cold sodium bicarbonate solution (10%), and
the aqueous layers were extracted twice with CH2Cl2. The
organic phases were filtered on a cotton plug and evaporated
under reduce pressure. The resulting crude products were then
preadsorbed on silica gel (10 g) in a 20 mL polypropylene
syringe, and elution was performed using EtOAc/hexanes, 1:9
(10 mL) to remove the unpolar byproducts. The desired more
polar secondary amines were recovered using EtOAc (20 mL)
in acceptable yields (50-70%) and directly used for the next
step (second level of reaction).
B-2-7. 3â-[(N-Cyclopropyl-N-butanoyl)aminomethyl]-3R-hy-
droxy-5R-androstan-17-one. IR (film): 3421 (OH), 1740
(CdO, ketone), 1629 (CdO, amide). 1H NMR (CDCl3): 0.78
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.97 (t, J ) 7.3 Hz, 3H,
CH3CH2), 0.80-2.20 (27H), 2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.2
Hz, 1H of CH2-16), 2.56 (t, J ) 7.5 Hz, 2H, CH2CO), 2.78 (m,
1H, CHN), 3.40 (s, 2H, CH2N-3â), 4.12 (s, OH). LRMS for
C27H44NO3 [M + H]+: 430.6 m/z. HPLC (CH3CN:H2O:CH3OH,
35:35:30): 97% of purity.
B-3-6. 3â-[(N-Cyclopentyl-N-cyclopropylcarbonyl)aminom-
ethyl]-3R-hydroxy-5R-androstan-17-one. IR (film): 3322 (OH),
1739 (CdO, ketone), 1610 (CdO, amide). 1H NMR (CDCl3):
0.77 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.80-2.15 (34H),
2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.24 (s,
2H, CH2N-3â), 4.43 (m, 1H, CHN), 5.56 (s, OH). LRMS for
C29H46NO3 [M + H]+: 456.6 m/z. HPLC (CH3CN:H2O:CH3OH,
40:25:35): 82% of purity.
B-4-1. 3â-[(N-2-Methoxymethyl-N-acetyl)aminomethyl]-3R-
hydroxy-5R-androstan-17-one. IR (film): 3364 (OH), 1735
(CdO, ketone), 1618 (CdO, amide). 1H NMR (CDCl3): 0.78
and 0.79 (2s of conformers A and B, 3H, CH3-19), 0.85 and
0.86 (2s of conformers A and B, 3H, CH3-18), 0.80-2.10 (22H),
2.13 and 2.16 (2s of conformers A and B, 3H, CH3CO), 2.42
(dd, J 1 ) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.35 (s, 3H,
CH3O), 3.20-3.80 (m, 6H, OCH2 and CH2CH2N and CH2N-
3â). LRMS for C25H42NO4 [M + H]+: 420.3 m/z. HPLC (CH3-
CN:H2O:CH3OH, 30:40:30): 97% of purity.
B-4-4. 3â-[(N-2′-Methoxyethyl-N-bromoacetyl)aminomethyl]-
3R-hydroxy-5R-androstan-17-one. IR (film): 3416 (OH), 1734
(CdO, ketone), 1636 (CdO, amide). 1H NMR (CDCl3): 0.79
(2s of conformers A and B, 3H, CH3-19), 0.85 and 0.86 (2s of
conformers A and B, 3H, CH3-18), 0.80-2.12 (24H), 2.42 (dd,
J 1 ) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.35 and 3.36 (2s,
conformers A and B, 3H), 3.20-3.80 (m, 6H OCH2 and CH2N
and CH2N-3â), 3.97 and 4.02 (2s, conformers A and B, 2H, CH2-
Br). LRMS for C25H42NO4 [M + H]+: 498.3 m/z. HPLC (CH3-
CN:H2O:CH3OH, 30:40:30): 95% of purity.
B-5-1. 3â-[(N-3′-Ethoxypropyl-N-acetyl)aminomethyl]-3R-
hydroxy-5R-androstan-17-one. IR (film): 3375 (OH), 1739
(CdO, ketone), 1623 (CdO, amide). 1H NMR (CDCl3): 0.78
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 1.20 (t, J ) 7.0 Hz, 3H,
CH3CH2), 0.90-2.15 (24H), 2.16 (s, 3H, CH3CO), 2.42 (dd, J 1
) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.33 (dapp of AB
system, 2H, CH2N-3â), 3.45 (m, 6H, CH2N and 2 × CH2O).
LRMS for C27H46NO4 [M + H]+: 448.5 m/z. HPLC (CH3CN:
H2O:CH3OH, 30:40:30): 97% of purity.
B-5-6. 3â-[(N-3′-Ethoxypropyl-N-cyclopropylcarbonyl)ami-
nomethyl]-3R-hydroxy-5R-androstan-17-one. IR (film): 3380
(OH), 1739 (CdO, ketone), 1614 (CdO, amide). 1H NMR
(CDCl3): 0.77 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 1.19 (t, J
) 7.0 Hz, 3H, CH3CH2), 0.80-2.15 (28H), 2.42 (dd, J 1 ) 8.7
Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.35 (qapp of AB system, 2H,
CH2N-3â), 3.46 (m, 4H, 2 x CH2O), 3.65 (m, 2H, CH2N), 4.68
(s, OH). LRMS for C29H48NO4 [M + H]+: 474.6 m/z. HPLC
(CH3CN:H2O:CH3OH, 35:35:30): 94% of purity.
3â-Am idom eth yl-3r-h yd r oxy-5r-a n d r osta n -17-on e (11).
Each of the secondary amines obtained above (eight for library
B and seven for library C) was split in 7 parts (0.05 mmol)
and dissolved in dry CH2Cl2 (1 mL). Piperidinomethyl poly-
styrene (40 mg, 0.15 mmol) and the appropriate acyl chloride
(0.06 mmol) were added to the appropriate reactors and stirred
for 2 h at room temperature. After the reaction was completed
(TLC monitoring), aminomethyl polystyrene (0.15 mmol) was
added to each reactor to scavenge the excess of unreacted acyl
chloride. After 1 h, the suspensions were filtered to remove
the resins and the solutions recovered were evaporated to give
the desired library B (56 members) and library C (49 members)
in good yields for both libraries (70-90%) and high average
purity for a random sampling of library B members, B-2-4,
B-2-7, B-3-6, B-4-1, B-4-4, B-5-1, B-5-6, B-6-1, B-6-6, B-6-7,
B-7-1, B-8-2, and B-8-4, and library C members, C-2-6, C-3-
7, C-4-3, C-4-5, C-5-1, C-6-1, C-6-2, and C-7-7 (Table 1). See
Figures 4 and 5 for the chemical structures of the compounds.
B-6-1. 3â-[(N-2′-Pentoxyethyl-N-acetyl)aminomethyl]-3R-hy-
droxy-5R-androstan-17-one. IR (film): 3378 (OH), 1740
(CdO, ketone), 1624 (CdO, amide). 1H NMR (CDCl3): 0.78
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.92 (t, J ) 7.3 Hz, 3H,
CH3CH2), 0.80-2.15 (27H), 2.15 (s, 3H, CH3CO), 2.42 (dd, J 1
) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.30-3.50 (m, 8H, 2 ×
CH2O and CH2N), 4.62 (OH). LRMS for C29H50NO4 [M + H]+:
476.3 m/z. HPLC (CH3CN:H2O:CH3OH, 35:30:35): 90% of
purity.

650 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Maltais et al.
B-6-6. 3â-[(N-2′-Pentoxyethyl-N-cyclopropylcarbonyl)ami-
nomethyl]-3R-hydroxy-5R-androstan-17-one. IR (film): 3361
(OH), 1739 (CdO, ketone), 1612 (CdO, amide). 1H NMR
(CDCl3): 0.77 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.92 (t, J
) 7.3 Hz, 3H, CH3CH2), 0.80-2.15 (33H), 2.42 (dd, J 1 ) 8.7
Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.34 (qapp of AB system, 2H,
CH2N-3â), 3.42 (m, 4H, 2 × CH2O), 3.64 (m, 2H, CH2N). LRMS
for C31H52NO4 [M + H]+: 502.4 m/z. HPLC (CH3CN:H2O:CH3-
OH, 35:30:35): 82% of purity.
B-6-7. 3â-[(N-2′-Pentoxyethyl-N-butanoyl)aminomethyl]-3R-
hydroxy-5R-androstan-17-one. IR (KBr): 3357 (OH), 1740
(CdO, ketone), 1618 (CdO, amide). 1H NMR (CDCl3): 0.77
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.93 and 0.96 (2t, J )
7.3 Hz, 6H, 2 × CH3CH2), 0.80-2.15 (30H), 2.36 (t, J ) 7.5
Hz, 2H, CH2CO), 2.42 (dd, J 1 ) 8.8 Hz, J 2 ) 19.5 Hz, 1H of
CH2-16), 3.30-3.50 (m, 8H, 2 × CH2O and CH2NCH2). LRMS
for C31H54NO4 [M + H]+: 504.6 m/z. HPLC (CH3CN:H2O:CH3-
OH, 45:20:35): 93% of purity.
B-7-1. 3â-[(N-Furan-2-ylmethyl-N-acetyl)-aminomethyl]-3R-
hydroxy-5R-androstan-17-one. IR (film): 3395 (OH), 1737
(CdO, ketone), 1631 (CdO, amide). 1H NMR (CDCl3): 0.77
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.80-2.15 (22H), 2.27
(s, 3H, CH3CO), 2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.1 Hz, 1H of
CH2-16), 3.37 (qapp of AB system, 2H, CH2N-3â), 4.02 (s, OH),
4.53 (s, 2H, CH2N), 6.22 (d, J ) 3.2 Hz, 1H, O-CHdCH), 6.34
(dd, J 1 )1.0, J 2 ) 3.0 Hz, 1H, CHdCH-O), 7.38 (d, J ) 1.1
Hz, 1H, O-CHdC). LRMS for C27H40NO4 [M + H]+: 442.6 m/z.
HPLC (CH3CN:H2O:CH3OH, 30:40:30): 99% of purity.
B-8-2. 3â-[(N-Allyl-N-dichloroacetyl)aminomethyl]-3R-hy-
droxy-5R-androstan-17-one. IR (film): 3456 (OH), 1737
(CdO, ketone), 1666 (CdO, amide). 1H NMR (CDCl3): 0.78
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.80-2.15 (22H), 2.42
(dd, J 1 ) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.42 (s, 2H,
CH2N-3â), 4.27 (d, J ) 2.0 Hz, 2H, CH2N), 5.17 (d, J ) 17.0
Hz, 1H of CH2dCH), 5.30 (d, J ) 10.5 Hz, 1H of CH2dCH),
CH aromatic), 7.64 (s, 1H, CH aromatic). LRMS for C33H44F6-
NO3 [M + H]+: 616.4 m/z. HPLC (CH3CN:H2O:CH3OH, 40:
20:40): 86% of purity.
C-4-5. 3â-[(N-3′,5′-Bis(trifluoromethyl)benzyl-N-2′-ethylhex-
anoyl)aminomethyl]-3R-hydroxy-5R-androstan-17-one. IR (film):
3340 (OH), 1738 (CdO, ketone), 1630 (CdO, amide). 1H NMR
(CDCl3): 0.76 (s, 3H, CH3-19), 0.83 (m, 6H, 2 × CH3CH2), 0.85
(s, 3H, CH3-18), 0.80-2.15 (30H), 2.32 (m, 1H, CHCO), 2.42
(dd, J 1 ) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.34 and 3.55
(2m, 2H, CH2N-3â), 4.88 (AB system, 2H, CH2N), 7.60 (s, 2H,
CH aromatic), 7.83 (s, 1H, CH aromatic). LRMS for C37H52F6-
NO3 [M + H]+: 672.6 m/z. HPLC (CH3CN:H2O:CH3OH, 45:
15:40): 97% of purity.
C-5-1. 3â-[(N-4′-Methylbenzyl-N-acetyl)aminomethyl]-3R-
hydroxy-5R-androstan-17-one. IR (film): 3405 (OH), 1738
(CdO, ketone), 1632 (CdO, amide). 1H NMR (CDCl3): 0.75
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.80-2.47 (22H), 2.15
(s, 3H, CH3CO), 2.35 (s, 3H, CH3-Ph), 2.42 (dd, J 1 ) 8.7 Hz,
J 2 ) 19.0 Hz, 1H of CH2-16), 3.37 (qapp of AB system, 2H,
CH2N-3â), 4.62 (s, 2H, CH2N), 7.02 (d, J ) 7.9 Hz, 2H, CH
aromatic), 7.18 (d, J ) 7.9 Hz, 2H, CH aromatic). LRMS for
C
30H44NO3 [M + H]+: 466.4 m/z. HPLC (CH3CN:H2O:CH3OH,
35:35:30): 98% of purity.
C-6-1. 3â-[(N-3′,5′-Bis(methoxy)benzyl-N-acetyl)aminomethyl]-
3R-hydroxy-5R-androstan-17-one. IR (film): 3405 (OH), 1737
(CdO, ketone), 1627 (CdO, amide). 1H NMR (CDCl3): 0.75
(s, 3H, CH3-19), 0.84 (s, 3H, CH3-18), 0.80-2.15 (21H), 2.14
(s, 3H, CH3CO), 2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.2 Hz, 1H of
CH2-16), 3.39 (qapp, 2H, CH2N-3â), 3.78 (s, 6H, 2 × OCH3), 4.17
(s, OH), 4.59 (s, 2H, CH2N), 6.25 (d, J ) 2.0 Hz, 2H, CH
aromatic), 6.36 (s, 1H, CH aromatic). LRMS for C31H46NO5 [M
+ H]+: 512.6 m/z. HPLC (CH3CN:H2O:CH3OH, 35:35:30): 98%
of purity.
C-6-2. 3â-[(N-3′,5′-Bis(methoxy)benzyl-N-propionyl)aminome-
thyl]-3R-hydroxy-5R-androstan-17-one. IR (film): 3416 (OH),
1738 (CdO, ketone), 1620 (CdO, amide). 1H NMR (CDCl3):
0.75 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 1.13 (t, J ) 7.4 Hz,
3H, CH3CH2), 0.80-2.15 (21H), 2.38 (m, 3H, CH3CH2CO and
1H of CH2-16), 3.40 (qapp of AB system, 2H, CH2N-3â), 3.78 (s,
6H, 2 × OCH3), 4.18 (s, OH), 4.60 (s, 2H, CH2N), 6.24 (d, J )
2.0 Hz, 2H, CH aromatic), 6.36 (d, J ) 1.9 Hz, 1H, CH
aromatic). LRMS for C32H48NO5 [M + H]+: 526.6 m/z. HPLC
(CH3CN:H2O:CH3OH, 35:35:30): 96% of purity.
5.81 (m, 1H, CH2dCH), 6.22 (s, 1H, CHCl2). LRMS for C25H38
-
Cl2NO3 [M + H]+: 470.4 m/z. HPLC (CH3CN:H2O:CH3OH, 35:
35:30): 92% of purity.
B-8-4. 3â-[(N-Allyl-N-dibromoacetyl)aminomethyl]-3R-hy-
droxy-5R-androstan-17-one. IR (film): 3396 (OH), 1735
(CdO, ketone), 1654 (CdO, amide). 1H NMR (CDCl3): 0.78
(s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.80-2.15 (22H), 2.43
(dd, J 1 ) 8.8 Hz, J 2 ) 19.1 Hz, 1H of CH2-16), 3.37 (s, 2H,
CH2N-3â), 3.84 (s, 2H, CH2Br), 4.17 (d, J ) 2.3 Hz, 2H, CH2N),
5.14 (d, 1H, J ) 17.2 Hz, 1H of CH2dCH), 5.26 (d, J ) 10.3
C-7-7. 3â-[(N-Adamantylmethyl-N-i-propylcarbonyl)ami-
nomethyl]-3R-hydroxy-5R-androstan-17-one. IR (film): 3394
(OH), 1738 (CdO, ketone), 1615 (CdO, amide). 1H NMR
(CDCl3): 0.78 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 1.13 (d, J
) 6.5 Hz, 6H, (CH3)2CH), 0.80-2.09 (36H), 2.42 (dd, J 1 ) 8.7
Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 2.95 (m, 1H, CHCO), 3.13 (s,
Hz, 1H of CH2dCH), 5.82 (m, 1H, CH2dCH). LRMS for C25H39
-
BrNO3 [M + H]+: 480.4 m/z (79Br) and 482.5 m/z (81Br). HPLC
(CH3CN:H2O:CH3OH, 30:40:30): 91% of purity.
C-2-6. 3â-[(N-2′-Phenylethyl-N-cyclopropylcarbonyl)ami-
nomethyl]-3R-hydroxy-5R-androstan-17-one. IR (film): 3390
(OH), 1738 (CdO, ketone), 1614 (CdO, amide). 1H NMR
(CDCl3): 0.77 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.80-2.15
(25H), 2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 2.94
(t, J ) 7.6 Hz, 2H, CH2Ph), 3.31 (qapp of AB system, 2H, CH2N-
3â), 3.76 (m, 2H, CH2N), 7.25 (m, 5H, CH aromatic). LRMS
for C32H46NO3 [M + H]+: 492.5 m/z. HPLC (CH3CN:H2O:CH3-
OH, 35:30:35): 99% of purity.
C-3-7. 3â-[(N-3′-Phenylpropyl-N-i-propylcarbonyl)aminome-
thyl]-3R-hydroxy-5R-androstan-17-one. IR (KBr): 3441 (OH),
1744 (CdO, ketone), 1608 (CdO, amide).1H NMR (CDCl3):
0.74 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 1.09 (d, J ) 6.7 Hz,
2H, CH2N-3â), 3.44 (s, 2H, CH2N), 4.71 (OH); LRMS for C35H56
-
NO3 [M + H]+: 538.7 m/z. HPLC (CH3CN:H2O:CH3OH, 45:
15:40): 76% of purity.
P r ep a r a tion of Libr a r y D Mem ber s (Meth od III). 3â-
Allyla m in om eth yl-3r-h yd r oxy-5r-a n d r osta n -17-on e (12).
To a solution of spiro-3(R)-oxirane-5R-androstan-17-one (4)30,46
(10.0 g, 34.4 mmol) in anhydrous ethanol was added allylamine
(4.5 mL, 68.8 mmol), and the solution was stirred overnight
at 60 °C. The resulting solution was evaporated under reduced
pressure and purified by flash chromatography using EtOAc/
hexanes (1:1) as the eluent to give compound 12 (9.8 g, 82%).
1
IR (KBr): 3500-3330 (NH and OH), 1737 (CdO, ketone). H
6H, (CH3)2CH), 0.80-2.15 (24H), 2.42 (dd, J 1 ) 8.6 Hz, J 2
)
NMR (CDCl3): 0.71 (s, 3H, CH3-19), 0.79 (s, 3H, CH3-18),
0.90-2.10 (24H), 2.36 (dd, J 1 ) 8.6 Hz, J 2 ) 18.9 Hz, 1H of
CH2-16), 2.42 (s, 2H, CH2N-3â), 3.21 (d, J ) 5.9 Hz, 2H,
19.1 Hz, 1H of CH2-16), 2.62 (t, 2H, J ) 7.2 Hz, CH2Ph), 2.60
(m, 1H, CHCO), 3.28 (s, 2H, CH2N), 3.33 (qapp of AB system,
2H, CH2N-3â), 7.25 (m, 5H, CH aromatic). LRMS for C33H50
-
CHCH2N), 5.03 (d, J ) 10.3 Hz, 1H, CH2dCH), 5.10 (dd, J 1 )
NO3 [M + H]+: 508.5 m/z. HPLC (CH3CN:H2O:CH3OH, 40:
25:35): 99% of purity.
1.3 Hz, J 2 ) 17.0 Hz, 1H, CH2dCH), 5.82 (m, 1H, CHdCH2).
13C NMR (CDCl3): 11.1, 13.7, 20.1, 21.6, 28.1, 30.7, 31.4, 31.6,
33.5, 34.9, 35.7, 35.9, 38.5, 40.5, 47.6, 51.3, 52.9, 54.1, 60.0,
69.8, 115.7, 136.7, 221.2. LRMS for C23H38NO2 [M + H]+: 360.2
m/z.
C-4-3. 3â-[(N-3′,5′-Bis(trifluoromethyl)benzyl-N-butanoyl)-
aminomethyl]-3R-hydroxy-5R-androstan-17-one. IR (KBr): 3440
(OH), 1739 (CdO, ketone), 1683 (CdO, amide). 1H NMR
(CDCl3): 0.76 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.91 (t, J
) 7.3 Hz, 3H, CH3CH2), 0.80-2.15 (24H), 2.27 (t, 2H, J ) 7.3
Hz, CH2CO), 2.43 (dd, J 1 ) 9.0 Hz, J 2 ) 19.4 Hz, 1H of CH2-
16), 3.42 (m, 2H, CH2N-3â), 4.86 (s, 2H, CH2N), 7.57 (s, 2H,
3R-Spir o-(3′-allyl-2′-oxo-oxazolidin -5′-yl)-5r-an dr ostan -
17-on e (14). To a solution of compound 12 (2.6 g, 7.2 mmol)
in anhydrous CH2Cl2 (200 mL) at 0 °C under an atmos-
phere of argon were added DIPEA (2.5 mL, 14.4 mmol) and

Type 3 17â-Hydroxysteroid Dehydrogenase Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 651
triphosgene (1.07 g, 3.6 mmol). The solution was stirred at
room temperature for 3 h. The resulting solution was diluted
with CH2Cl2 (300 mL) and washed with a water/ice mixture
(1 L). The organic layer was filtered on a cotton plug and
evaporated under reduced pressure. The resulting crude
product was purified by flash chromatography using EtOAc/
hexanes (4:6) as the eluent to give compound 14 (2.52 g, 91%).
IR (KBr): 1734 (CdO, ketone and carbamate). 1H NMR
(CDCl3): 0.81 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.90-2.15
(21H), 2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.16
(s, 2H, CH2N-3â), 3.84 (d, J ) 5.5 Hz, 2H, CH2N), 5.20 (d, J )
14.8 Hz, 1H, CH2dCH), 5.21 (d, J ) 12.2 Hz, 1H, CH2dCH),
5.73 (m, 1H, CH2dCH). 13C NMR (CDCl3): 11.4, 13.8, 20.2,
21.7, 27.8, 30.5, 31.4, 32.8, 33.8, 35.0, 35.3, 35.8, 39.4, 40.8,
46.6, 47.7, 51.3, 53.9, 55.9, 78.8, 118.4, 132.1, 157.2, 221.2.
LRMS for C24H36NO3 [M + H]+: 386.1 m/z.
was evaporated to dryness under reduced pressure and im-
mediately dissolved in methanol (170 mL) containing K2CO3
(2.19 g) and then vigorously stirred for 3 h. The resulting
solution was poured in a 10% ammonium chloride solution (500
mL) and extracted twice with EtOAc (2 × 150 mL). The
resulting crude product was purified by flash chromatography
using EtOAc/hexanes (1:1) as the eluent to give compound 17
(1.38 g, 65%). IR (film): 1734 (CdO, ketone and carbamate).
1H NMR (CDCl3): 0.80 (s, 3H, CH3-19), 0.83 (s, 3H, CH3-18),
0.90-2.15 (21H), 2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.1 Hz, 1H of
CH2-16), 2.54 (d, J ) 2.1 Hz, 1H, CH2(O)CH), 2.78 (t, 1H, CH2-
(O)CH), 3.05 (m, 2H, CH2N), 3.22 and 3.29 (2d of AB system,
2H, CH2N-3â), 3.80 (m, 1H, CH2(O)CH).13C NMR (CDCl3):
11.3, 13.7, 20.1, 21.7, 27.8, 30.5, 31.4, 32.6, 33.8, 34.9, 35.3,
35.8, 39.2, 40.8, 44.2, 46.0, 47.7, 50.3, 51.2, 53.8, 57.4, 79.8,
157.4, 221.0. LRMS for C24H36NO4 [M + H]+: 402.2 m/z.
Syn th esis of In ter m ed ia te Ca r ba m a te 18. To compound
17 (5 × 200 mg, 0.50 mmol) in dry CH3CN (1 mL) were added
the appropriate secondary amine (1.0 mmol) and anhydrous
lithium perchlorate (64 mg, 0.60 mmol). The solutions were
heated at 60 °C overnight in sealed vials. Methylisocyanate
polystyrene (600 mg, 1.0 mmol/g) and CH2Cl2 (2 mL) were
added to each reaction vessel and stirred for 2 h at room
temperature. The five resins were removed by filtration, and
the resulting filtrates were diluted with CH2Cl2 (15 mL) and
washed with water (25 mL). The resulting solutions were
filtered on a cotton plug and then evaporated under reduced
pressure to give the five desired â-amino-alcohols 18 in good
yields (60-80%) and high TLC purities.
Syn th esis of Com p ou n d 19. Each of the five â-amino
alcohols 18 obtained above were split in 5 parts (0.05 mmol)
and dissolved with EtOAc (1 mL). Piperidinomethyl polysty-
rene (40 mg, 0.15 mmol) and DMAP-PS (15 mg, 0.02 mmol)
were added to each of the 25 reactors. The acyl chlorides (0.50
mmol) were added to the appropriate reactors and shaken for
48 h at 72 °C under an atmosphere of argon. After the reaction
was completed (TLC monitoring), aminomethyl polystyrene
resin (0.55 mmol, 200 mg) and CH2Cl2 (1 mL) were added to
each reactor to scavenge the excess of unreacted acyl chloride.
After 1 h, the suspensions were filtered and the recovered
solutions were evaporated to give the crude products contain-
ing a trace of carboxylic acid, a byproduct coming from the
hydrolysis of the acyl chloride during the reaction. Therefore,
the library members were dissolved with EtOAc (5 mL) and
washed with a 10% NaHCO3 solution. Finally, the products
were filtered on sillica gel (3 g) and eluted first with hexanes
(10 mL) and then EtOAc (20-30 mL) to recover the desired
library members in moderate yields (20-40%) and purity
higher than 80% (by TLC and 1H NMR analysis) for a sampling
of library D members D-1-1, D-2-2, D-3-3, D-4-4, and D-5-5
(Table 1). See Figure 6 for the chemical structures of com-
pounds.
3R-Sp ir o-(3′-a d a m a n tylm eth yl-2′-oxo-oxa zolid in -5′-yl)-
5r-a n d r osta n -17-on e (15). To a solution spiro-3(R)-oxirane-
5R-androstan-17-one (4)30,40 (1.0 g, 3.44 mmol) in ethanol (99%
in benzene) was added 1-adamantane-methylamine (1.2 mL,
6.88 mmol), and the solution was stirred overnight at 60 °C.
The resulting solution was evaporated under reduced pressure
and purified by flash chromatography using EtOAc/hexanes
(1:1) as the eluent to give compound 13 (1.1 g, 70%). To a
solution of compound 13 (1.0 g, 2.1 mmol) in anhydrous CH2-
Cl2 (30 mL) at 0 °C under an atmosphere of argon were added
DIPEA (458 µL, 2.6 mmol) and triphosgene (227 mg, 0.75
mmol). The solution was stirred at room temperature for 30
min. The resulting solution was diluted with CH2Cl2 (150 mL)
and washed with a water/ice mixture (1 L). The organic layer
was then filtered on a cotton plug and evaporated under
reduced pressure. The resulting crude product was puri-
fied by flash chromatography using CH2Cl2/MeOH (95:5) as
the eluent to give compound 15 (0.82 g, 79%). IR (film): 1739
1
(CdO, ketone and carbamate). H NMR (CDCl3): 0.81 (s, 3H,
CH3-19), 0.85 (s, 3H, CH3-18), 0.80-2.15 (37H), 2.42 (dd, J 1
)
8.7 Hz, J 2 ) 19.0 Hz, 1H of CH2-16), 2.85 (s, 2H, CH2N), 3.30
(s, 2H, CH2N-3â). 13C NMR (CDCl3): 11.4, 13.8, 20.2, 21.7,
27.9, 28.2 (3×), 30.6, 31.5, 32.7, 33.9, 35.0, 35.4, 35.8, 36.8 (5×),
39.3, 40.8 (3×), 47.7, 51.3, 53.9, 57.1, 60.4, 78.3, 158.6, 221.1.
LRMS for C32H48NO3 [M + H]+: 494.6 m/z. HPLC (MeOH:
H2O: 72: 28 + AcNH4 (20 mM)): 98% of purity.
3R-Spir o-(3′-(2′′,3′′-dih ydr oxypr opyl)-2′-oxo-oxazolidin -
5′-yl)-5r-a n d r osta n -17-on e (16). To a solution of compound
14 (2.49 g, 6.5 mmol) in tert-butyl alcohol/THF/H2O (10:3:1)
(70 mL) were added NMO (908 mg, 7.8 mmol) and osmium
tetroxide (50 mg, 0.2 mmol), and the solution was stirred for
2 h at room temperature. The mixture was then diluted with
EtOAc (250 mL) and poured in a 10% (w/v) sodium bisulfite
solution (750 mL). The organic layer was washed with water
(750 mL) and brine (300 mL), dried with MgSO4, filtered, and
evaporated under reduce pressure. The resulting crude product
was purified by flash chromatography using CH2Cl2/MeOH (9:
1) as the eluent to give compound 16 (2.43 g, 90%). IR (KBr):
3404 (OH), 1734 (CdO, ketone and carbamate). 1H NMR
(CDCl3): 0.79 (s, 3H, CH3-19), 0.83 (s, 3H, CH3-18), 0.80-2.15
(23H), 2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.33
(m, 4H, CH2N-3â and CH2N), 3.57 (m, 2H, CH2OH), 3.84 (m,
1H, CHOH). 13C NMR (CDCl3): 11.3, 13.7, 20.2, 21.7, 27.7,
30.5, 31.4, 32.7, 33.8, 34.9, 35.3, 35.8, 39.2, 40.8, 46.4, 47.7,
D-1-1. 3R-Spiro-{3′-[3′′-diethylamino-2′′-(cyclopropylcarbo-
nyl)propyl]-2′-oxo-oxazolidin-5′-yl}-5R-androstan-17-one. IR
(film): 1742 (CdO, ketone, ester and carbamate). 1H NMR
(CDCl3): 0.81 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.80-2.15
(25H), 1.02 (t, J ) 6.7 Hz, 6H, 2 × CH3CH2), 2.42 (dd, J 1
)
8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 2.57 (m, 7H, 3x CH2N
and CHCO), 3.18-3.37 and 3.47 (3m, 4H, CH2-3â and CH2-
NCOO), 5.15 (broad s, 1H, CHOCO). LRMS for C32H51N2O5
[M + H]+: 543.5 m/z.
51.2, 53.8, 58.0, 63.7, 70.2, 79.8, 158.6, 221.3. LRMS for C24H38
-
NO5 [M + H]+: 420.4 m/z.
D-2-2. 3R-Spiro-{3′-[3′′-dipropylamino-2′′-(cyclohexylcarbo-
nyl)propyl]-2′-oxo-oxazolidin-5′-yl}-5R-androstan-17-one. IR
(film): 1742 (CdO, ketone, ester and carbamate). 1H NMR
(CDCl3): 0.81 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.86 (t, J
) 7.3 Hz, 6H, 2 × CH3CH2), 0.80-2.15 (35H), 2.40 (m, 8H, 3
× CH2N, CHCO and 1H of CH2-16), 3.13, 3.35 and 3.60 (3m,
4H, CH2N-3â and CH2NCOO), 5.15 (broad s, 1H, CHOCO).
LRMS for C37H61N2O5 [M + H]+: 613.2 m/z.
3R-Sp ir o-(3′-(2′′,3′′-ep oxyp r op yl)-2′-oxo-oxa zolid in -5′-
yl)-5r-a n d r osta n -17-on e (17). To a solution of diol 16 (2.21
g, 5.3 mmol) in anhydrous CH2Cl2 (40 mL) under an atmos-
phere of argon were added trimethylorthoacetate (808 µL, 6.3
mmol) and pyridinium p-toluenesulfonate (PPTS) (67 mg, 0.3
mmol). The solution was stirred for 15 min at room temper-
ature and then immediately evaporated under reduced pres-
sure, followed by 1 min under a mechanical pump. The
resulting crude mixture was dissolved in anhydrous CH2Cl2
(10 mL), TEA (53 mg, 0.5 mmol, 66 µL) and acetyl bromide
(780 mg, 63 mmol) were then added, and the mixture was
stirred for 30 min under an atmosphere of argon. The solution
D-3-3. 3R-Spiro-{3′-[3′′-N-pyrolidine -2′′-(3′′′-benzoyloxy)-
propyl]-2′-oxo-oxazolidin-5′-yl}-5R-androstan-17-one. IR (film):
1
1738 (CdO, ketone, ester and carbamate). H NMR (CDCl3):
0.77 (s, 3H, CH3-19), 0.84 (s, 3H, CH3-18), 0.80-2.15 (25H),
2.42 (dd, J 1 ) 8.7 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 2.69 (broad

652 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3
Maltais et al.
s, 4H, 2 × CH2N), 2.88 (m, 2H, NCH2-CHO), 3.25 and 3.38
(2d of AB system, 2H, CH2N-3â), 3.50 and 3.70 (2m, 2H, CH2-
NCOO), 5.47 (broad s, 1H, CHOCO), 7.45 (m, 2H, CH
aromatic), 7.58 (m, 1H, CH aromatic), 8.03 (d, J ) 7.5 Hz, 2H,
CH aromatic). LRMS for C35H49N2O5 [M + H]+: 577.4 m/z.
D-4-4. 3R-Spiro-{3′-[3′′-N-piperidine- 2′′-(3′′′-cyclopentyl-
propionyloxy)propyl]-2′-oxo-oxazolidin-5′-yl) }-5R-androstan-
17-one. IR (film): 1739 (CdO, ketone, ester and carbamate).
1H NMR (CDCl3): 0.81 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18),
0.80-2.15 (38H), 2.32 (t, J ) 7.7 Hz, 2H, CH2CO), 2.42 (m,
7H, 3 × CH2N and 1H of CH2-16), 3.17, 3.35 and 3.55 (3m,
4H, CH2N-3â and CH2NCOO), 5.20 (broad s, 1H, CHOCO).
LRMS for C37H59N2O5 [M + H]+: 611.1 m/z.
the removal of the dioxolane group (acetone/HCl 3M, 1:1) to
give the desired 3â-[N-1-adamantylmethyl-N-butyl)amino-
methyl]-3R-hydroxy-5R-androstan-17-one (20). IR (film): 3523
(OH), 1737 (CdO, ketone). 1H NMR (CDCl3): 0.74 (s, 3H, CH3-
19), 0.84 (s, 3H, CH3-18), 0.89 (t, J ) 7.2 Hz, 3H, CH3CH2),
0.80-2.15 (41H), 2.17 (s, 2H, CH2N), 2.34 (s, 2H, CH2N), 2.42
(m, 3H, CH2N and 1H of CH2-16), 4.09 (s, OH). 13C NMR
(CDCl3): 11.2, 13.8, 14.1, 20.2, 20.5, 21.7, 28.4 (4×), 29.8, 30.7,
31.6, 34.3 (2×), 34.7, 35.0, 35.6, 35.8, 37.0 (2×), 41.1, 41.2, 41.7-
(4×), 47.8, 51.4, 54.3, 59.3, 69.0, 69.7, 71.2, 221.4. LRMS for
C
35H58NO2 [M + H]+: 524.4 m/z. HRMS calcd for C35H58NO2
[M + H]+, 524.4468; found, 524.4473.
In h ibition of Typ e 3 17â-HSD (Scr een in g of Libr a r ies
A-D) a n d IC50 Deter m in a tion . P r ep a r a tion of th e En -
zym a tic Sou r ce. The expression vectors encoding for type 3
17â-HSD were transfected into HEK-293 cells using the
calcium phosphate procedure.2 Cells were then sonicated in
50 mM sodium phosphate buffer (pH 7.4), containing 20%
glycerol and 1 mM of ethylenediaminetetraacetic acid (EDTA),
and centrifugated at 10 000g for 1 h to remove the mitochon-
dria, plasma membranes, and cell fragments. The supernatant
was further centrifugated at 100 000g to separate the microso-
mal fraction, which was used as the source of type 3 17â-HSD
activity for the enzymatic assay.
D-5-5. 3R-Spiro-{3′-[3′′-N-morpholino-2′′-(4′′′-tert-butyl-ben-
zoyloxy)propyl]-2′-oxo-oxazolidin-5′-yl)}-5R-androstan-17-
1
one. IR (film): 1738 (CdO, ketone, ester and carbamate). H
NMR (CDCl3): 0.77 (s, 3H, CH3-19), 0.83 (s, 3H, CH3-18), 1.33
(s, 9H, (CH3)3C), 0.80-2.15 (21H), 2.42 (dd, J 1 ) 8.7 Hz, J 2
)
19.2 Hz, 1H of CH2-16), 2.55 (m, 6H, 3 × CH2N), 3.20, 3.40
and 3.70 (3m, 4H, CH2N-3â and CH2NCOO), 3.66 (broad s,
4H, 2 × CH2O), 5.41 (broad s, 1H, CHOCO), 7.46 (d, J ) 7.7
Hz, 2H, CH aromatic), 7.93 (d, J ) 8.3 Hz, 2H, CH aromatic).
LRMS for C39H57N2O6 [M + H]+: 649.5 m/z.
P u r ifica tion of Selected Typ e 3 17â-HSD In h ibitor s for
Libr a r ies A-D. Before, to be used for IC50 determination and
cell proliferative assay (Table 2), the inhibitors A-4-11, B-1-
6, C-7-3, and D-5-4 were synthesized again, purified by flash
chromatography (SiO2), and characterized.
En zym a tic Assa y. The inhibition test was carried out at
37 °C in 1 mL of 50 mM sodium phosphate buffer (pH 7.4),
containing 20% glycerol and 1 mM of EDTA, 5 mM of cofactor
(NADPH), 0.05 µM [4-14C] 4-androstene-3-17-dione (New
England Nuclear, Boston, MA), and the indicated concentra-
tion of compounds to be tested. The reaction was stopped after
2 h by adding 2 mL of diethyl ether containing 10 µM of
unlabeled 4-androstene-3-17-dione (∆4-dione) and T. The
metabolites were extracted twice with 2 mL of diethyl ether,
evaporated, and then dissolved in CH2Cl2 before being applied
on silica gel 60 TLCs. TLC plates were developed in a mixture
of toluene and acetone (4:1). Substrate [14C] ∆4-dione and
metabolite [14C] T were identified by comparison with reference
steroids and revealed by autoradiography and then quantified
using the PhosphorImager (Molecular Dynamics, Sunnyvale,
CA). The percentage of transformation and the percentage of
inhibition were calculated from eqs 1 and 2, respectively.
A-4-11. 3â-[(N-Octyl-N-cyclopropylcarbonyl)aminomethyl]-
3R-hydroxy-5R-androstan-17-one. IR (film): 3364 (OH), 1734
1
(CdO, ketone), 1608 (CdO, amide). H NMR (CDCl3): 0.77 (s,
3H, CH3-19), 0.84 (s, 3H, CH3-18), 0.88 (t, J ) 6.2 Hz, 3H,
CH3CH2), 0.95-2.20 (39H), 2.42 (dd, J 1 ) 8.5 Hz, J 2 ) 18.8
Hz, 1H of CH2-16), 3.32 (qapp of AB system, 2H, CH2N-3â), 3.47
(m, 2H, CH2CH2N), LRMS for C32H54NO3 [M + H]+: 500.7 m/z.
HRMS calcd for
500.4090. HPLC (CH3CN:H2O:CH3OH, 35:20:45): 98% of
purity.
C
32H54NO3 [M + H]+, 500.4103; found,
B-1-6. 3â-[(N-Cyclohexylmethyl-N-cyclopropylcarbonyl)ami-
nomethyl]-3R-hydroxy-5R-androstan-17-one. IR (film): 3420
(OH), 1740 (CdO, ketone), 1616 (CdO, amide). 1H NMR
(CDCl3): 0.78 (s, 3H, CH3-19), 0.84 (s, 3H, CH3-18), 0.80-2.20
(37H), 2.42 (dd, J 1 ) 8.6 Hz, J 2 ) 19.2 Hz, 1H of CH2-16), 3.35
% transformation ) ([14C]-T/([14C]-T + [14C]-∆4-dione)) ×
100 (1)
(m, 4H, CH2N-3â and CH2N), 4.65 (s, OH). LRMS for C31H50
-
NO3 [M + H]+: 484.7 m/z. HRMS calcd for C31H50NO3 [M +
H]+, 484.3791; found, 484.3781. HPLC (CH3CN:H2O:CH3OH,
40:20:40): 99% of purity.
% inhibition ) [(% transformation without inhibitor -
% transformation with inhibitor)/
% transformation without inhibitor] × 100 (2)
C-7-3. 3â-[(N-Adamantylmethyl-N-butanoyl)aminomethyl]-
3R-hydroxy-5R-androstan-17-one. IR (film): 3392 (OH), 1734
The IC50 value, the concentration of inhibitor that gives 50%
of enzymatic inhibition, was calculated by a computer (DE50
program, CHUL Research Center, Que´bec, Canada).
P r olifer a t ive Act ivit y on Sh ion ogi (AR+) Ma m m a r y
Cells. Assays on the proliferation of AR+ Shionogi mammary
carcinoma cells were carried out according to the procedure
previously described by Sam et al.41 Two concentrations of
tested compound (0.1 and 1 µM) were used in the assay. The
results were expressed as the percentage (%) of proliferation,
as compared to the stimulation induced by 0.3 nM of the
androgen DHT (100%).
1
(CdO, ketone), 1616 (CdO, amide). H NMR (CDCl3): 0.78 (s,
3H, CH3-19), 0.85 (s, 3H, CH3-18), 0.96 (t, J ) 7.3 Hz, 3H,
CH3CH2), 1.00-2.20 (38H), 2.40 (m, 3H, CH2CH2CO and 1H
of CH2-16), 3.11 (s, 2H, CH2N), 3.47 (s, 2H, CH2N-3â), 4.82 (s,
OH). LRMS for C35H56NO3 [M + H]+: 538.6 m/z. HRMS calcd
for C35H56NO3 [M + H]+, 538.4260; found, 484.4265. HPLC
(H2O containing 0.1% TFA): 95% of purity.
D-5-4. 3R-Spiro-{3′-[3′′-N-morpholino-2′′-(3′′′-cyclopentyl-
propionyloxy)propyl]-2′-oxo-oxazolidin-5′-yl)}-5R-androstan-17-
one. IR (film): 1738 (CdO, ketone, ester and carbamate). 1H
NMR (CDCl3): 0.81 (s, 3H, CH3-19), 0.85 (s, 3H, CH3-18),
0.80-2.20 (33H), 2.33 (t, J ) 7.5 Hz, 2H, CH2CO), 2.46 (m,
7H, 3 × CH2N and 1H of CH2-16), 3.18, 3.35 and 3.56 (3m,
4H, CH2N-3â and CHCH2N), 3.66 (broad s, 4H, CH2O), 5.19
(m, 1H, CHOCO). LRMS for C36H57N2O6 [M + H]+: 613.2.4
m/z. HPLC (CH3CN:H2O:CH3OH, 55:25:20): 90% of purity.
HRMS calcd for C36H58N2O6 [M + H]+, 613.4217; found,
613.4246.
Syn th esis of Com p ou n d 20 (Rep or ted in Ta ble 2).
Compound 9 was submitted to aminolysis using 1-adaman-
tylmethylamine and then to an acylation reaction with valeryl
chloride under the conditions reported above for method II,
except that the dioxolane group was not hydrolyzed (Scheme
2). The tertiary amide 11 with a C17-dioxolane group was then
reduced using a solution of BH3 (1.0 M) in THF followed by
Ack n ow led gm en t. We thank the Medical Research
Council of Canada (MRC) and Le Fonds de la Recherche
en Sante´ du Que´bec (FRSQ) for their financial support,
the Division of Medicinal Chemistry (Oncology and
Molecular Endocrinology Research Center) for providing
chemical and biological facilities, Dr. Fernand Labrie
for Shionogi cell culture assay, and Dr. Agne`s Coquet
and Mrs. Lison Nadeau for HPLC and LRMS analysis.
Also, we warmly thank M. Patrick Be´langer for LCMS
analysis and useful discussions. Finally, we are very
grateful to Mrs. Marie Be´rube´ for the preparation of
library C members.

Type 3 17â-Hydroxysteroid Dehydrogenase Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 3 653
(22) Simard, J .; Luthy, I.; Guay, J .; Be´langer, A.; Labrie, F. Char-
acteristics of interaction of the antiandrogen flutamide with the
androgen receptor in various target tissues. J . Mol. Cell.
Endocrinol. 1986, 44, 261-270.
Refer en ces
(1) Inano, H.; Tamaoki, B.-I. Testicular 17â-hydroxysteroid dehy-
drogenase: molecular properties and reaction mechanism. Ste-
roids 1986, 48, 1-26.
(23) Palmberg, C.; Koivisto, K. L.; Tammela, T. L.; Kallioniemi, O.
P.; Visakorpi, T. Androgen receptor gene amplification at
primary progression predicts response to combined androgen
blockade as second line therapy for advanced prostate cancer.
J . Urol. 2000, 164, 1992-1995.
(2) Luu-The, V.; Zhang, Y.; Poirier, D.; Labrie, F. Characteristics
of human types 1, 2 and 3 17â-hydroxysteroid dehydrogenase
activities: Oxidation/reduction and inhibition. J . Steroid Bio-
chem. Mol. Biol. 1995, 55, 581-587.
(3) Labrie, F.; Luu-The, V.; Lin, S.-X.; Labrie, C.; Simard, J .; Breton,
R.; Be´langer, A. The key role of 17â-hydrosteroid dehydrogenase
in sex steroid biology. Steroids 1997, 62, 148-158.
(24) Poirier, D.; Labrie, F.; Luu-The, V. Inhibition de la 17â-
hydroxyse´troide de´shydroge´nase (17â-HSD) type
3 par des
(4) Petoketo, H.; Luu-The, V.; Simard, J .; Adamski, J . 17beta-
hydroxysteroid dehydrogenase (HSD)/17-ketosteroid reductase
(KSR) family; nomenclature and main characteristics of the 17
HSD/KSR enzymes. J . Mol. Endocrinol. 1999, 23, 1-11.
(5) Labrie, F. Intracrinology. Mol. Cell. Endocrinol. 1991, 78, C113-
C118.
(6) El-Alfy, M.; Luu-The, V.; Huang, X.-F.; Berger, L.; Labrie, F.;
Pelletier, G. Localization of type 5 17â-hydroxysteroid dehydro-
genase, 3â-hydroxysteroid dehydrogenase, and androgen recep-
tor in the human prostate by in situ hybridization and immu-
nocytochemistry. J . Endocrinol. 1999, 140, 1-11.
(7) Dufort, I.; Rheault, P.; Fang, X.; Soucy, P.; Luu-The, V. Char-
acteristics of a highly labile human type 5 17â-hydroxysteroid
dehydrogenase. Endocrinology 1999, 140, 568-574.
(8) Rittmaster, R. S.; Antonian, L.; New, M. I.; Stoner, E. Effect of
finasteride on adrenal steroidogenesis in men. J . Androl. 1994,
15, 298-301.
(9) Hirosumi, J .; Nakayama, O.; Fagan, T.; Sawada, K.; Chida, N.;
Inami, M.; Takahashi, S.; Kojo, H.; Notsu, Y.; Okuhara, M.
FK143, a novel nonsteroidal inhibitor of steroid 5R-reductase.
1. in vitro effects on human and animal prostatic enzymes. J .
Steroid Biochem. Mol. Biol. 1995, 52, 357-363.
(10) Russel, D. W.; Wilson, J . D. Steroid 5R-reductase: two genes/
two enzymes. Annu. Rev. Biochem. 1994, 63, 25-61.
(11) Poirier, D.; Ciobanu, L. C.; Maltais, R. Steroid sulfatase inhibi-
tors. Exp. Opin. Ther. Pat. 1999, 9, 1083-1099.
(12) Labrie, F.; Cusan, L.; Labrie, C.; Simard, J .; Luu-The, V.;
Diamond, P.; Gomez, F.; Candas, B. History of LHRH agonist
and combination therapy in prostate cancer. Endocr.-Relat.
Cancer 1996, 3, 243-278.
(13) Van Weerden, W.; van Steenbrugge, G.; van Kreuningen, A.;
Moerings, E.; De J ong, F.; Schroder, F. Effects of low testosterone
levels and of adrenal androgens on growth of prostate tumor
models in nude mice. J . Steroid Biochem. Mol. Biol. 1990, 20,
903-907.
(14) Liu, X.; Wiley, H.; Meikle, A. Androgens regulate proliferation
of human prostate cancer cells in culture by increasing trans-
forming growth factor-alpha (TGF-alpha) and epidermal growth
factor (EGF)/TGF-alpha receptor. J . Clin. Endocrinol. 1993, 77,
1472-1478.
de´rive´s ste´roidiens: EÄ tude exploratoire. Medecine-Sciences 1995,
11, 24.
(25) Tche´dam-Ngatcha, B.; Luu-The, V.; Poirier, D. Androsterone 3â-
substituted derivatives as inhibitors of type 3 17â-hydroxysteroid
dehydrogenase. Bioorg. Med. Chem. Lett. 2000, 10, 2533-2536.
(26) Ciobanu, L. C.; Maltais, R.; Poirier, D. The sulfamate functional
group as a new anchor for solid-phase organic synthesis. Org.
Lett. 2000, 2, 445-448.
(27) Tremblay, M. R.; Simard, J .; Poirier, D. Parallel solid-phase
synthesis of a model library of 7R-alkylamide estradiol deriva-
tives as potential estrogen receptor antagonists. Bioorg. Med.
Chem. Lett. 1999, 9, 2827-2832.
(28) Maltais, R.; Luu-The, V.; Poirier, D. Parallel solid-phase syn-
thesis of 3â-peptido-3R-hydroxy-5R-androstan-17-one derivatives
for inhibition of type
3 17â-hydroxysteroid dehydrogenase.
Bioorg. Med. Chem. 2001, 9, 3101-3111.
(29) Maltais, R.; Be´rube´, M.; Marion, O.; Labrecque, R.; Poirier, D.
Efficient coupling and solid-phase synthesis of steroidal ketone
derivatives using polymer-bound glycerol. Tetrahedron Lett.
2000, 41, 1691-1694.
(30) Maltais, R.; Poirier, D. A solution-phase combinatorial parallel
synthesis of 3â-amido-3R-hydroxy-5R-androstane-17-one. Tet-
rahedron Lett. 1998, 39, 4151-4154.
(31) Leznoff, C. C.; Wong, J . Y. The use of polymer supports in organic
synthesis. VII. Polymer-bound 1,3-diols as monoblocking agents
of symmetrical dialdehydes. Can. J . Chem. 1973, 51, 3824-3829.
(32) Leznoff, C. C.; Greenberg, S. The use of polymer supports in
organic synthesis. III. Selective chemical reactions on one
aldehyde group of symmetrical dialdehydes. Can. J . Chem. 1976,
54, 3756-3764.
(33) Chini, M.; Crotti, P.; Macchia, F. J . Regioalterning selectivity
in the metal salt catalyzed aminolysis of styrene oxide. J . Org.
Chem. 1991, 56, 5939-5942.
(34) Military Standard: Sampling Procedures and Tables for Inspec-
tion by Attributes; Department of Defense (United States of
America): Washington D.C., 1963; Vol. 1, p 67.
(35) Novabiochem. The Combinatorial Chemistry Catalog, 1999.
(36) Eckert, H.; Foster, B. Triphosgene, a crystalline phosgene
substitute. Angew. Chem., Int. Ed. Engl. 1987, 26, 894.
(37) Kolb, H. C.; Sharpless, K. B. A simplified procedure for the
stereospecific transformation of 1,2-diols into epoxides. Tetra-
hedron. 1992, 48, 10515-10530.
(38) Ward, D. E.; Rhee, C. K. A simple method for the microscale
preparation of mosher’s acid chloride. Tetrahedron Lett. 1991,
32, 7165-7166.
(15) Marchetti, B.; Labrie, M.; Poulin, R.; Labrie, F. Castration levels
of plasma testosterone have potent stimulatory effects on
androgen-sensitive parameters in the rat prostate. J . Steroid
Biochem. 1988, 31, 411-419.
(16) Seidenfeld, J .; Samson, D. J .; Hasselblad, V.; Aronson, N.;
Albertsen, P. C.; Bennett, C. L.; Wilt, T. J . Single-therapy
androgen suppression in men with advanced prostate cancer:
a systematic review and meta-analysis. Ann. Intern. Med. 2000,
132, 566-577.
(17) Santen, R. J . Clinical Review 37: Endocrine treatment of
prostate cancer. J . Clin. Endocrinol. Metab. 1992, 75, 685-689.
(18) Gheiler, E.; Tiguert, R. Current concepts in androgen deprivation
therapy: is there a “best” endocrine treatment? World J . Urol.
2000, 18, 190-193.
(19) Simard, J .; Singh, S.; Labrie, F. Comparison of in vitro effects
of the pure antiandrogen OH-flutamide, casodex, and nilutamide
on androgen-sensitive parameters. J . Urol. 1997, 49, 580-586.
(20) Oefelein, M. G.; Cornum, R. Failure to achieve castrate levels
of testosterone during luteinizing hormone releasing hormone
agonist therapy: the case for monitoring serum testosterone and
a treatment decision algorithm. J . Urol. 2000, 164, 726-729.
(21) Oefelein, M. G. Serum testosterone-based luteinizing hormone-
releasing hormone agonist redosing schedule for chronic andro-
gen ablation: a phase I assessment. Urology 1999, 54, 694-
699.
(39) Tokunaga, M.; Larrow, J . F.; Kakiuchi, F.; J acobsen, E. N.
Asymmetric catalysis with water: efficient kinetic resolution of
terminal epoxides by means of catalytic hydrolysis. Science 1997,
277, 936-938.
(40) Maltais, R.; Tremblay, M. R.; Poirier, D. Solid-Phase synthesis
of hydroxysteroid derivatives using the diethylsilyloxy linker.
J . Comb. Chem. 2000, 2, 604-614.
(41) Sam, K. M.; Labrie, F.; Poirier, D. N-Butyl-N-methyl-11-(3′-
hydroxy-21′,17′-carbolactone-19′-nor-17′R-pregna-1′,3′,5′(10′)-
trien-7′R-yl)-undecanamide: an inhibitor of type 2 17â-hydroxy-
steroid dehydrogenase that does not have oestrogenic or andro-
genic activity. Eur. J . Med. Chem. 2000, 35, 217-225.
J M010286Y