Chemical synthesis, NMR analysis and evaluation on a cancer xenograft model (HL-60) of the aminosteroid derivative RM-133

Article abstract of DOI:10.1016/j.steroids.2014.01.008

The aminosteroid derivative RM-133 has been reported to be a promising pro-apoptotic agent showing activity on various cancer cell lines. Following the development of solid-phase synthesis that generated a series of libraries of aminosteroid derivatives, we now report the development of a convenient liquid phase chemical synthesis of RM-133, the most promising candidate, in order to obtain sufficient quantities to proceed with the first preclinical assays. A simple and convergent six-step synthesis was designed and allowed the preparation of a gram-quantity scale of RM-133. This aminosteroid derivative was also fully characterized by NMR experiments which revealed an interesting mixture of conformers. Finally, the in vivo potency of RM-133 was evaluated on a xenograft model in nude mice with HL-60 tumors, which has resulted in the blocking of tumor progression by 57%.

Full text of DOI:10.1016/j.steroids.2014.01.008

Steroids 82 (2014) 68–76
Contents lists available at ScienceDirect
Steroids
journal homepage: www.elsevier.com/locate/steroids
Chemical synthesis, NMR analysis and evaluation on a cancer xenograft
model (HL-60) of the aminosteroid derivative RM-133
⇑
René Maltais, Audrey Hospital, Audrey Delhomme, Jenny Roy, Donald Poirier
Laboratory of Medicinal Chemistry, CHU de Québec – Research Center (CHUL, T4), and Faculty of Medicine (Université Laval), Québec, Québec G1V 4G2, Canada
a r t i c l e i n f o
a b s t r a c t
Article history:
The aminosteroid derivative RM-133 has been reported to be a promising pro-apoptotic agent showing
activity on various cancer cell lines. Following the development of solid-phase synthesis that generated
a series of libraries of aminosteroid derivatives, we now report the development of a convenient liquid
phase chemical synthesis of RM-133, the most promising candidate, in order to obtain sufficient quanti-
ties to proceed with the first preclinical assays. A simple and convergent six-step synthesis was designed
and allowed the preparation of a gram-quantity scale of RM-133. This aminosteroid derivative was also
fully characterized by NMR experiments which revealed an interesting mixture of conformers. Finally,
the in vivo potency of RM-133 was evaluated on a xenograft model in nude mice with HL-60 tumors,
which has resulted in the blocking of tumor progression by 57%.
Received 19 November 2013
Received in revised form 9 January 2014
Accepted 20 January 2014
Available online 30 January 2014
Keywords:
Synthesis
Cancer
Leukemia
Xenograft
NMR
Ó 2014 Elsevier Inc. All rights reserved.
1
. Introduction
The efficiency of most of the anticancer drugs currently used in
replaced by a quinoline to increase the hydrophilicity of the com-
pound. These modifications of E-37P led to the optimized aminos-
teroid derivative RM-133, which showed a stronger cytotoxic
activity on a panel of breast (MCF-7, Shionogi), prostate (LNCaP)
and leukemia (HL-60) cancer cell lines [14]. RM-133 was thus se-
lected as our lead compound to perform preclinical studies
in vivo. However, our initial solid-phase chemical route to obtain
RM-133 was not adapted to provide the multi-gram quantity of
compound required to carry out the in vivo studies. Herein, we re-
ported a new and efficient convergent chemical synthesis of RM-
133, a detailed NMR analysis of this novel aminosteroid derivative
and the results of a first in vivo preclinical study showing promis-
ing anticancer activities.
clinical settings is limited by their inherent toxicities on normal
cells and by the progressive apparition of drug resistance during
treatment [1,2]. The development of new anticancer agents with
greater selectivity that work through different molecular mecha-
nisms of action is strongly needed to respond to these two impor-
tant issues in cancer chemotherapy [3]. Aminosteroid derivatives
represent a new family of molecules with promising anticancer
activity on various cancer cells lines and showing a selectivity of
action on normal cells [4–7]. Over the last few years, we have syn-
thesized a large series of aminosteroid derivatives by parallel
solid-phase synthesis and classic organic synthesis from which
important structure–activity relationships have emerged (Fig. 1)
2
. Materials and methods
[
8–11]. Among the libraries synthesized, we have identified one
member, called E-37P, as one of the most potent aminosteroid
derivatives [5]. We then performed different chemical modifica-
tions on E-37P’s core by classic organic synthesis to improve its
pharmacokinetic properties. Notably, an ethinyl group was intro-
2
.1. Chemistry
Androsterone, epi-androsterone and chemical reagents were
purchased from Sigma–Aldrich Canada Ltd. (Oakville, ON, Canada).
The usual solvents were obtained from Fisher Scientific (Montréal,
QC, Canada) and were used as received. Anhydrous dichlorometh-
ane (DCM), tetrahydrofuran (THF) and dimethylformamide (DMF)
were obtained from Sigma–Aldrich. Ethyl acetate (EtOAc), hexanes
and methanol (MeOH) were purchased from Fisher Scientific. Thin-
layer chromatography (TLC) and flash-column chromatography
were performed on 0.20-mm silica gel 60 F254 plates (E. Merck;
Darmstadt, Germany) and with 230–400 mesh ASTM silica gel 60
duced at position C-17a to avoid glucuronidation of 17b-OH, a
strategy used with success for other steroid scaffolds [12–14].
The naphthalene group on the 2b-piperazine moiety was also
⇑
Corresponding author. Address: Laboratory of Medicinal Chemistry, CHU de
Québec – Research Center (CHUL, T4-42), 2705 Laurier Boulevard, Québec, Québec
G1V 4G2, Canada. Tel.: +1 418 654 2296; fax: +1 418 654 2298.
E-mail address: donald.poirier@crchul.ulaval.ca (D. Poirier).
http://dx.doi.org/10.1016/j.steroids.2014.01.008
0
039-128X/Ó 2014 Elsevier Inc. All rights reserved.

R. Maltais et al. / Steroids 82 (2014) 68–76
69
R
1
H
4 3 2
R R R
H
H
HO
H
R
1
=
- OH (androstane) or CH(OH)CH
3
(pregnane)
Steroid scaffold
Piperazine
A
R2
=
N
O
N
O
O
H
N
Diversity elements
of the libraries
R
=
N
3
N
Best amino acids
retained from 25
natural and unnatural
amino acids
(Pro)
(Phe)
(TIC)
O
R
4
=
O
O
S
2
R
R
R
R
N
H
Different functionalities
used as capping groups
(
amides)
(sulfonamides)
(alkylamines)
(ureas)
B
C
Screening of 300 compounds on different cell lines
and selection of candidates showing highest
anticancer activity
O
O
OH
N
N
H
N
H
H
HO
H
st
E-37P
(1 generation)
Structural modifications at various positions
Y
X
O
N
O
R
1
/ R
2
: H or CH
3
/ H
OR
R
2
1
Optimisation of
E-37P properties
Z
or Ethinyl
N
H
N
O
R: H or SO NH
2
2
H
H
R
H
X, Y, Z : N or C
D
Selection of one candidate regarding
cytotoxicity, selectivity and bioavailability
O
O
OH
N
N
N
H
N
H
H
HO
H
(2nd generation)
RM-133
Fig. 1. Representation of the steps behind the development of the aminosteroid RM-133. (A) Solid-phase synthesis of aminosteroid libraries provided crucial structure–
activity relationship (SAR) data where androstane (R ) piperazine (R ), proline (R ) and amide functionality (R ) were identified as important elements for anticancer activity.
B) Screening of the library members on different cancer cell lines identified E-37P as one of the most active compounds synthesized. (C) Structural modifications of E-37P
improved the biological properties and in vivo stability. (D) RM-133 emerged as the most promising aminosteroid for in vivo studies.
1
2
3
4
(
(
Silicyle, Québec, QC, Canada), respectively. Infrared spectra (IR)
spectrometer (Billerica, MA, USA). The chemical shifts (d) were ex-
pressed in ppm and referenced to chloroform (7.26 and 77.0 ppm),
acetone (2.06 and 29.2 ppm) or methanol (3.31 ppm and
49.0 ppm), respectively for H and C NMR. Low-resolution mass
spectra (LRMS) were recorded on a Shimadzu apparatus (Kyoto,
were recorded on a Horizon MB 3000 ABB FTIR spectrometer (Qué-
bec, QC, Canada) and the significant band reported in cm . Nucle-
ar magnetic resonance (NMR) spectra were recorded at 400 MHz
ꢀ1
1
13
1
13
for H and 100.6 MHz for C on a Bruker Avance 400 digital

7
0
R. Maltais et al. / Steroids 82 (2014) 68–76
Japan) equipped with a turbo ion-spray source. High-resolution
mass spectra (HRMS) were provided by Pierre Audet at the Laval
University Chemistry Department (Québec, QC, Canada).
and 2: 4.0 g, 10.6 mmol; assay 3: 3.4 g, 9.0 mmol) in anhydrous
THF (assays 1 and 2: 235 mL; assay 3: 200 mL). The solution was
allowed to return at room temperature and stirred for 7 h. The
reaction was stopped by the addition of ice/water and the mixture
2.1.1. Preparation of 5
a-androst-2-en-17-one (1)
2 4
extracted with EtOAc, washed with brine, dried with Na SO , fil-
To a solution of epi-androsterone (assay 1: 5.0 g, 17.2 mmol; as-
say 2: 23.0 g, 79.2 mmol; assay 3: 20.0 g, 68.9 mmol) in anhydrous
DCM (assay 1: 150 mL; assay 2: 690 mL; assay 3: 600 mL) at
78 °C under an argon atmosphere was added diethylaminosulfur
trifluoride (DAST) (assay 1: 2.7 mL, 20.7 mmol; assay 2: 12.5 mL,
5.0 mmol; assay 3: 10.8 mL, 82.6 mmol) and the solution was stir-
red for 1 h at ꢀ78 °C. The solution was then directly evaporated
with silica gel and purified by flash chromatography with EtOAc/
hexanes (1:9) as eluent to give the alkene 1 (yield for assay 1:
tered and evaporated under reduced pressure to give 4.3 g assay
1; 4.9 g assay 2 and 4.0 g assay 3 of corresponding trimethylsilyl-
ethinyl intermediate compound. To the combined compounds
from assays 1–3 (13.0 g, 27.4 mmol) in MeOH (750 mL) was added
ꢀ
K
2
CO
room temperature and then filtered to remove the excess of
CO . The methanol filtrate was concentrated under reduced
3
(37.5 g, 271.0 mmol). The solution was stirred overnight at
9
K
2
3
pressure to ꢂ30 mL, water (1 L) added and the solution extracted
with DCM (3 ꢁ 250 mL) and EtOAc (2 ꢁ 250 mL). The combined or-
1
.96 g, 39%; assay 2: 5.32 g, 23%; assay 3: 4.22 g, 21%) with 3
oro-5 -androstan-17-one as by-product (ratio 75:25 of 1 vs.
-fluoro compound). The ratio of C –C alkene vs. C –C alkene
isomer is less than 90:10 (calculated using the CH-4 signal at
.3 ppm). IR (ATR) ) d: 0.78 (s, 18-
: 1736 (C@O). 1H NMR (CDCl
CH ), 0.87 (s, 19-CH ), 0.9–2.0 (residual CH and CH ), 2.06 (m,
-CH), 2.44 (dd, J = 8.8 Hz and J = 19.2 Hz, 16b-CH), 5.59 (m,
-CH and 3-CH). C NMR (CDCl ) d: 11.7, 13.7, 20.2, 21.8, 28.4,
0.2, 30.6, 31.6, 34.7, 35.1, 35.8, 39.7, 41.4, 47.7, 51.4, 54.1,
a
-flu-
2 4
ganic phase was dried with Na SO , filtered and evaporated under
a
reduced pressure to give 10.3 g (93%) of crude compound 6. The
compound was used without further purification for the next step.
3a
2
3
3
4
IR (KBr)
m
: 3395 and 3310 (OH and NH), 2102 (C„CH, very weak).
H NMR (acetone-d ) d: 0.84 (s, 19-CH ), 1.01 (s, 18-CH ), 0.70–
2.00 (residual CH and CH ), 2.15 (m, 1H), 2.37 (m, 2 -H), 2.45
and 2.58 (2 m, 2 ꢁ CH N), 2.83 (m, 2 ꢁ CH N), 2.93 (s, C„CH),
4.01 (m, 3b-CH). C NMR (CD OD) d: 13.4, 14.7, 21.9, 24.0, 29.0,
32.8, 34.0, 34.3, 34.4, 37.1, 37.2, 39.8, 40.5, 46.6 (2ꢁ), 48.0 (2ꢁ),
41 2 2
51.6, 51.7, 56.6, 66.4, 66.8, 74.7, 80.3, 88.9. LRMS for C25H N O
1
5
m
3
6
3
3
3
3
2
2
a
1
2
3
1
6
a
1
2
2
2
1
3
13
3
3
+
25.7, 125.8, 221.5. LRMS for C19
H
29O [M+H] 273.1 m/z.
+
[
M+H] 401.3 m/z.
2.1.2. (2a, 3a)-2,3-epoxiandrostan-17-one (2)
To a solution of alkene 1 containing 25% of 3
a
-fluoro-5a-andro-
2.1.5. tert-butyl 1-(quinolin-2-ylcarbonyl)- -prolinate (7)
L
stan-17-one (5.04 g) in DCM (300 mL) was added m-chloroperben-
zoic acid (m-CPBA) 77% (3.78 g, 16.9 mmol) at 0 °C under an
atmosphere of argon. The solution was stirred at 4 °C for 24 h.
The solution was then diluted with DCM (500 mL) and washed
with an aqueous sodium bicarbonate solution (10%). The organic
layer was dried with sodium sulfate and evaporated under reduced
pressure. The crude epoxide was purified by flash chromatography
To a solution of quinaldic acid (10.0 g, 57.7 mmol) in anhydrous
DMF (200 mL) was added benzotriazole-1-yl-oxy-tris-pyrrolidino-
phosphonium hexafluorophosphate (PyBOP) (36.0 g, 69.2 mmol)
and N-hydroxybenzotriazole (HOBt) (9.43 g, 69.2 mmol). The solu-
tion was stirred for 5 min and proline-OtBu (12.0 g, 57.8 mmol)
was added before the addition of DIPEA (40 mL, 0.22 mol). The
resulting solution was stirred at room temperature under an argon
atmosphere overnight. The reaction was poured into water (1 L)
and extracted three times with EtOAc, washed with brine, dried
(
(
1
hexanes/EtOAc: 8:2) to yield pure compound 2 (2.73 g, 66%). IR
1
ATR)
9-CH
dd, J
ide), 3.16 (broad s, CH of epoxide). C NMR (CDCl
m
3
: 1736 (C@O). H NMR (CDCl
), 0.60–1.96 (residual CH and CH
= 8.7 Hz, J = 18.8 Hz, 16b-H), 3.11 (t, J = 5.8 Hz, CH of epox-
3
) d: 0.79 (s, 18-CH
3
), 0.84 (s,
-H), 2.43
2
), 2.05 (m, 16
a
2 4
with Na SO , filtered and evaporated under reduced pressure.
The crude compound was purified by flash chromatography with
(
1
2
1
3
3
) d: 12.9, 13.7,
0.1, 21.7, 28.1, 29.0, 30.5, 31.5, 33.8, 35.2, 35.8, 36.2, 38.2, 47.6,
DCM/MeOH (97:3) as eluent to give the desired compound 7
1
2
5
(13.1 g, 63%). H NMR (CDCl
3
) d: 1.18 and 1.51 (2s, C(CH
), 2.34 (m, 1H of
), 4.61 and 5.41 (2dd,
), 7.60 (m, CH-12), 7.74 (m,
3 3
) , 1.98
+
0.9, 51.2, 52.3, 53.8, 221.2. LRMS for C19
H
29
O
2
[M+H] 289.3 m/z.
(m, NCH
NCHCH ), 3.90 and 4.08 (2m, NCH
= 3.9 Hz and J = 8.5 Hz, NCHCH
2 2 2
CH ), 2.06 and 2.16 (2m, 1H of NCHCH
2
2
2
.1.3. (2b, 3
To epoxide 2 (9.16 g, 31.5 mmol) were added piperazine
107.5 g, 1.58 mol) and water (40 mL). The solution was heated
a
)-3-hydroxy-2-(piperazin-1-yl)androstan-17-one (3)
J
1
2
2
CH-13), 7.82 and 7.85 (2d, J = 8.5 Hz, CH-11), 7.96 and 8.13 (2d,
J = 8.5 Hz, CH-8), 8.10 (m, CH-14), 8.23 and 8.24 (2d, J = 8.5 Hz,
(
1
3
at 130 °C over a period of 48 h. The hot resulting mixture was
poured into water (2 L) and extracted with DCM (2 ꢁ 500 mL).
CH-9). C NMR (CDCl
3
) d: 22.2 (25.3), 27.7 (28.0) (3ꢁ), 32.1
(29.0), 48.5 (49.8), 62.6 (60.9), 80.9 (81.3), 121.2 (121.0), 127.5
(127.6), 127.8 (127.7), 128.2, 129.6 (129.8), 129.7 (129.9), 136.7,
2 4
The organic phase was dried with Na SO , filtered and evaporated
under reduced pressure. The crude compound 3 (12.0 g) was used
without further purification for the next step. IR (ATR) : 3380 (OH
) d: 0.86 and 0.87 (2s, 18-CH
), 0.73–1.95 (residual CH and CH ), 2.06 (m, 16 -CH),
.45, 2.64 and 2.91 (3 m, 4ꢁ CH N and 2 -CH and 16b-CH), 3.85
) d: 13.8, 17.2, 20.6, 21.7, 28.1, 30.4,
146.1 (146.5), 152.8 (153.4), 165.6 (166.3), 172.1 (171.4). LRMS
+
m
for C19
H
23
N
2
O
3
[M+H] 327.2 m/z.
1
and NH), 1730 (C@O). H NMR (CDCl
and 19-CH
3
3
3
2
a
2.1.6. 1-(quinolin-2-ylcarbonyl)-L-proline (8)
2
2
a
To a solution of compound 7 (13.0 g, 35.8 mmol) was added a
solution of TFA/DCM (95:5) (150 mL). The solution was stirred at
room temperature for 3 h. The resulting solution was evaporated
under reduced pressure. The resulting brown oil was co-evapo-
1
3
(
m, 3b-CH). C NMR (CDCl
3
3
4
3
1.6, 32.6, 34.6, 35.0, 35.8 (2ꢁ), 38.4, 46.5 (4ꢁ CH
2
of piperazine),
+
7.9, 51.3, 56.1, 63.4, 65.0, 221.2. LRMS for C23
H
39
N
2
O
2
[M+H]
75.3 m/z.
2
rated with a mixture of THF/H O five times until obtaining a solid.
The solid was dried at 50 °C overnight to give compound 8 (13.2 g,
1
2
.1.4. (2b,3
a
,5
a
,17
a
)-2-(piperazin-1-yl)pregn-20-yne-3,17-diol (6)
99%) as light brown solid. H NMR (CDCl
2.10 and 2.22 (2 m, 1H of NCHCH ), 2.40 (m, 1H of NCHCH
3.92 (m, NCH ), 4.68 and 5.28 (2dd, = 4.3 Hz,
NCHCH ), 7.66 (m, CH-12), 7.79 (m, CH-13), 7.84 and 8.01 (2d,
3
) d: 2.03 (m, NCH
2
CH
2
),
),
To a solution of trimethylsilylacetylene (assays 1 and 2: 6.0 mL,
2.3 mmol; assay 3: 5.13 mL, 36.1 mmol) in anhydrous diethyl-
2
2
4
2
J
1
2
J = 8.6 Hz,
ether (assays 1 and 2: 235 mL; assay 3: 200 mL) under an argon
atmosphere was dropwise added MeLi (1.6 M in THF) (assays 1
and 2: 19.9 mL, 31.9 mmol; assay 3: 16.9 mL, 27.1 mmol) at 0 °C.
The mixture was stirred at room temperature for 1 h and cooled
again at 0 °C before the addition of crude compound 3 (assays 1
2
J = 8.6 Hz, CH-8), 7.94 and 7.98 (2d, J = 8.5 Hz, CH-11), 8.11 and
8.12 (2d, J = 8.5 Hz, CH-14), 8.39 and 8.44 (2d, J = 8.5 Hz, CH-9).
3
OD) d: 23.0 (26.3), 32.7 (30.2), 49.6 (51.0), 63.4
(61.4), 121.8 (121.4), 128.8 (129.0), 129.3 (129.2), 129.8, 130.4
1
3
C NMR (CD

R. Maltais et al. / Steroids 82 (2014) 68–76
71
(
1
130.3), 131.2 (131.5), 138.4 (138.8), 147.4 (147.8), 153.6 (154.5),
weighed, anesthetized with isoflurane, and killed by cervical
dislocation.
+
67.7 (168.5), 176.7 (175.4). LRMS for C15
15 2 3
H N O [M+H] 271.1
m/z.
2
.3. Statistical analysis
2
1
1
.1.7. {4-[(2b,3a,5a,17a)-3,17-dihydroxypregn-20-yn-2-yl]piperazin-
-yl}[(2S)-1-(quinolin-2-ylcarbonyl)pyrrolidin-2-yl]methanone (RM-
33)
Data are presented as means ± SEM. Statistical significance was
determined according to the multiple-range Duncan-Kramer test
15]. P-values that were less than 0.05 were considered as statisti-
cally significant.
To a solution of compound 8 (11.4 g, 30.7 mmol) in anhydrous
[
DMF (430 mL) under an argon atmosphere was added 2-(6-
chloro-1H-benzotriazole-1-yl)-1,1,3,3-tetramethylaminium hexa-
fluorophosphate (HCTU) (12.7 g, 30.7 mmol). The solution was
stirred for 5 min before the addition of compound 6 (10.3 g,
3
. Results and discussion
2
5.6 mmol) and DIPEA (26.7 mL, 153.4 mmol). The resulting solu-
tion was stirred overnight at room temperature. The reaction mix-
ture was poured into water (2 L), extracted three times with EtOAc,
3.1. Chemical synthesis of RM-133
2 4
washed with brine and dried with Na SO . The organic phase was
Our synthetic efforts in the preparation of the aminosteroid
filtered and evaporated under reduced pressure. The crude com-
pound was purified by two successive chromatographies (first:
DCM/MeOH (95:5); second: EtOAc/hexanes (7:3)) to give RM-133
derivative RM-133 were first directed to providing a more conve-
nient chemical route to the key intermediate 2b-piperazinyl-
androsterone (3) (Scheme 1). Although this compound was ob-
tained by a regioselective aminolysis of epoxide 2, which was eas-
ily obtained by a stereoselective epoxidation of the C2,C3-alkene 1,
an important ratio of undesired C3,C4-isomer of alkene 1, generally
varying from 10% to 25% depending of the methods used [9], was
however obtained. For example, the tosylation of epi-androsterone
(
5.75 g, 35%). IR (KBr)
m
: 3420 and 3306 (OH), 2101 (C„CH, very
weak), 1643 (CON). H NMR (CDCl ) d: 0.77 (m, CH-5), 0.82 and
-19), 0.84 and 0.88 (2s, CH -18), 0.80–1.85 (m, CH
-11, CH-8, CH-14, CH-9, CH -4, CH -12),
1
3
0
7
1
.86 (2s, CH
3
3
2
-
, CH -6, CH
.90–2.70 (3 -CH
2
2
0
-15, CH
2
2
2
0
00
2
, 4 -CH
2
, CH-2
a
, CH
2
-16), 2.57 and 2.58 (2s,
0
00
C„CH), 3.10–4.30 (m, 2ꢁ CH
= 3.3 Hz and J
dd, J = 4.1 Hz and J
2
-2 , CH-3b and CH
2
-5 ), 5.12 (dd,
(
epi-ADT) followed by elimination is a straightforward two-step
00
J
1
2
= 8.0 Hz, CH-2 (NCHCO) of one rotamer), 5.88
methodology to obtain 1, albeit in a large proportion (24%) of
C3,C4-isomer [10]. Unfortunately, these two isomeric olefins, as
well as their subsequent epoxides, were unseparable from each
other, especially in gram-scale. In fact, obtaining the pure C2,C3-al-
kene 1, or a mixture with less than 10% of C3,C4-isomer, was only
possible by using the classic method that necessitates six chemical
reactions (acetylation, bromination, reduction, elimination, hydro-
lysis and oxidation) from dihydrotestosterone and several chro-
matographic steps. In an alternative strategy, using a mixture of
C2,C3- and C3,C4-epoxides [16], the less hindered C2,C3-epoxide
00
(
7
7
1
2
= 8.5 Hz, CH-2 , (NCHCO) of one rotamer),
00
00
00
.59 (m, CH-12 ), 7.74 (m, CH-13 ), 7.85 (tapp, J = 6.6 Hz, CH-11 ),
00
.97 and 8.02 (2d, J = 8.5 Hz, CH-8 ), 8.00 and 8.12 (2d, J = 8.1 Hz,
00
00 13
CH-14 ), 8.22 and 8.24 (2d, J = 8.7 Hz, CH-9 ). C NMR (CDCl
3
) d:
1
CH
2.8 (CH
3
-18), 17.0 (CH
3
-19), 20.9 (21.2) (CH
2
-11), 22.4 (25.3)
00
2
-4 ), 22.9 (CH-15), 27.9 (28.0) (CH
2
-6), 31.0 (CH
2
-7), 28.9
00
(
(
(
4
(
31.5) (CH
34.4) (CH
2 2 2
-3 ), 32.5 (32.8) (CH -1), 32.6 (32.7) (CH -12), 34.3
2
-4), 35.6 (35.7) (C-10), 35.9 (36.0) (CH-8), 38.2 (38.3),
0
CH-9), 38.8 (38.9) (CH
6.9 (C-13), 47.7, 48.0, 48.4 (CH
CH-14), 55.3 (55.4) (CH-5), 57.4 (59.0) (CH-2 ), 63.6 (63.7) CH-3,
2
-16), 42.0, 42.5, 45.3 and 46.1 (CH
2
-2 ),
0
00
2 2
-1 ), 48.2 (49.8) (CH -5 ), 50.2
2
was selectively opened (aminolysis) in presence of ethylene gly-
00
col and piperazine under controlled temperature (130 °C) [10]. This
methodology has a limiting effect on the yield of the reaction by
producing an incomplete reaction and, consequently, an acceptable
yield was only obtained with successive repetition of the aminoly-
sis step, which was very time consuming at the preparation and
purification levels.
6
1
6
1
4.5 (64.8) (CH-2), 73.6 (73.7) (C-20), 79.6 (C-17), 87.7 (C-21),
000
000
20.9 (121.6) (CH-3 ), 127.5 (127.6) (CH-7 ), 127.5 (127.7) (CH-
0
00
000
000
000
), 128.1 (128.2) (C-5 ), 129.1 (CH-9 ), 129.6 (129.7) (CH-8 ),
000
000
000
36.7 (CH-4 ), 145.7 (146.3) (C-10 ), 153.3 (154.0) (C-2 ), 166.1
000
00
(
53 4 4
166.5) (C-1 ), 170.0 (170.3) (C-1 ). HRMS: calcd for C40H N O :
+
6
52.3989, found [M+H] 652.3995. HPLC purity of 99.5% (Retention
time = 15.8 min). HPLC purity of RM-133 was determined using a
RP-C18 column (Altima, HP, C18-AQ, 4.6 ꢁ 250 mm, 5 m), a gradi-
ent of solvents, from MeOH/H O (70:30) to 100% MeOH over
0 min, and a PDA detector at a maximum wavelength absorption
of 190 nm.
It became obvious that we had to develop another method to
obtain a sufficiently pure C2,C3-alkene 1 in order to efficiently pro-
vide multi-gram quantities of key intermediate 3. Recently, we
demonstrated that a mixture of C2,C3- and C3,C4-alkenes was ob-
tained as side-products when ADT and epi-ADT were treated with
diethylaminosulfur trifluoride (DAST), a fluorination agent [17,18].
Indeed, the use of 1.2 equivalent of DAST at ꢀ78 °C in DCM gave an
interesting ratio of alkenes (C2,C3:C3,C4) for ADT (95:5) and epi-
ADT (96:4) with a higher yield of alkene for ADT (57%) than for
epi-ADT (30%) [11]. However, since this dehydration reaction is
the first step of the synthesis and considering the low cost of
epi-ADT compared to ADT (five times less costly), we found advan-
tageous to use epi-ADT as a starting material rather than ADT, even
with an expected two-times lower yield of 1. Thus, the reaction
with epi-ADT and 1.2 eq of DAST at ꢀ78 °C in DCM gave the alkene
1 with a high C2,C3:C3,C4-ratio (96:4) and a moderate yield of 30%.
l
2
3
2.2. In vivo anticancer activity
Two million human leukemia (HL)-60 cells were inoculated
subcutaneously (s.c.) in 0.1 mL of growth medium (RPMI 1640
Sigma, Saint Louis, MO, USA) containing -glutamine (2 nM) and
antibiotics (100 IU penicillin/mL, 100 g streptomycin/mL) and
(
L
l
supplemented with 10% fetal bovine serum (FBS)) + 30% Matrigel
on both flanks of each mouse through a 2.5-cm long 25G needle.
After approximately 1 week, the mice were randomly assigned to
two groups according to tumor size. 1500
l
g of RM-133 (60 mg/
The major product of the reaction, the 3a-fluoro-5a-androstan-17-
kg) was suspended in 0.1 mL of propylene glycol:ethanol (92:8)
and then injected s.c. every 2 days (group 2, n = 10). Animals in
group 1 (n = 4, control) received 0.1 mL of the vehicle alone. The
one, which was difficult to separate from 1 at this stage by flash
chromatography, was easily removed by chromatography after
the next reaction (epoxidation) step. Overall, the DAST procedure
proved to be more appropriate than the methods for alcohol dehy-
dratation that were previously used [9], which always gave a high-
er amount (>10%) of undesired steroidal C3,C4-alkene. We also
size of the tumors was measured by two perpendicular diameters
2
(
L and W) and tumor area (mm ) was calculated using the formula
L/2 ꢁ W/2 ꢁ
p. After 39 days of treatment, the animals were

7
2
R. Maltais et al. / Steroids 82 (2014) 68–76
18
O
1
2
O
O
19
11
1
7
1
13
a
b
9
1
6
2
14
15
1
0
8
O
HO
3
5
7
4
6
1
2
Epiandrosterone (3β-OH)
Androsterone (3α-OH)
c
N
O
O
O
O
N
N
d
HN
e
N
N
N
O
O
ROUTE A
HO
HO
OH
N
Si
4
3
N
N
e, f
ROUTE B
HO
5
OH
f
N
HN
O
N
O
OH
N
g
N
HO
N
6
HO
RM-133
Scheme 1. Two chemical routes to synthesize RM-133: the first one (Route A; dashed arrows) and the optimized (Route B; full arrows). Reagents and conditions: (a) DAST,
DCM, ꢀ78 °C; (b) m-CPBA, DCM, 0 °C; (c) piperazine, water, 130 °C; (d) 1-(quinolin-2-ylcarbonyl)- -proline TFA salt (8), PyBOP, HOBt, DIPEA, DMF, rt; (e) trimethylsilylethinyl,
MeLi, ether/THF; (f) MeOH, K CO (10%), rt; (g) 1-(quinolin-2-ylcarbonyl)- -proline TFA salt (8), HCTU, DIPEA, DMF, rt.
L
2
3
L
favoured the one-step DAST procedure over the six-step classical
procedure that we previously used starting from DHT [8].
The subsequent stereoselective epoxidation of the olefin 1 (con-
taining 4% of C3,C4-isomer) was carried out with m-CPBA in DCM
which first appeared less attractive to us than route A because of
the potential interference of the unprotected 2b-piperazine group
with the trimethylsilylethinyl lithium reagent, worked remarkably
well. The ethinylation proceeded in quantitative yield and subse-
quent hydrolysis of the trimethylsilyl group with potassium car-
bonate in MeOH provided compound 6 in quantitative yield
without further purification needed. The subsequent acylation
reaction with the activated anhydride generated from carboxylic
acid 8 provided RM-133 in an acceptable yield, using HCTU as a
coupling reagent. The use of HCTU greatly facilitates the isolation
of the final compound when compared to PyBOP where phospho-
nium by-products contaminated the final compound and required
multiple chromatographic steps to remove this impurity.
giving good yields of the desired 2,3a-epoxide 2. Since we had ac-
cess to a sufficiently pure epoxide, it was now possible to perform
the aminolysis using piperazine in water at 130 °C for 48 h to drive
the reaction at completion. The 2b-piperazinyl-ADT (3) was ob-
tained in an excellent yield after a simple washing with water.
At this point of the synthesis, two different routes (A and B,
Scheme 1) were investigated to obtain RM-133. In route A, we first
added the 1-(quinolin-2-ylcarbonyl)-L-proline moiety by an acyla-
tion reaction of the corresponding anhydride with 2b-piperazinyl-
ADT (3) to generate 4. The quinoline–proline moiety 8 was previ-
ously prepared by coupling the 2-quinaldic acid with proline-t-
butylester using PyBOP coupling reagent followed by TFA depro-
tection of the t-butylester (Scheme 2). In a second step, we then
proceeded with the ethinylation of the 17-ketone of 4 using the
lithiated trimethylsilyl ethinyl to generate 5. Unfortunately, the
ethinylation reaction was found to be unselective over the amide
bounds and the resulting mixture of compounds was difficult to
separate, resulting in a low isolated yield of RM-133. We next
decided to investigate an alternative synthetic route. Route B,
3.2. NMR characterization of RM-133
The NMR spectra of the aminosteroid derivative RM-133 is par-
ticularly interesting as well as complex, and a deeper investigation
was required to understand this molecule’s behaviour. In fact, RM-
1
33 contains three characteristic moieties: a 17
a-ethinyl-5a-
androstane-3 ,17b-diol core (right part), piperazine group
a
a
located at position 2b of steroid (middle part) and a proline quin-
oline side chain on the piperazine (left part). These three different
O
OH
O
O
3
CF COO
O
O
H
H
N
a
O
O
N
b
OH
N
N
+
O
N
N
7
8
Scheme 2. Preparation of the side chain required for the final coupling step. Reagents and conditions: (a) PyBOP, HOBt, DIPEA, DMF, rt; (b) TFA/DCM (95:5), rt.

R. Maltais et al. / Steroids 82 (2014) 68–76
73
chemical entities on the same molecule greatly complexify the
NMR characterization and are responsible for a duplication of sev-
ond we concentrated our efforts on the presence of two conformers
depending on the solvent.
For the purposes of discussion, we started our analysis with the
17a-ethinyl-5a-androstane-3a,17b-diol moiety (right part of RM-
133). This steroid nucleus is well known and some of its NMR sig-
nals can be assigned from the data reported in literature [19–21],
1
13
eral signals in both H (Fig. 2) and C NMR (Fig. 3) spectra. The
shape and chemical shift of signals are also influenced at different
levels by the solvent used for NMR analysis. In the first part we ad-
1
13
dressed the assignation of H and C NMR signals and in the sec-
Fig. 2. 1H NMR spectra of aminosteroid derivative RM-133 in chloroform-d
1
4 6 6
(A), methanol-d (B), acetone-d (C) and dimethylsulfoxide-d (D). Most of the signals are
influenced differently according to the solvent.
Fig. 3. 13C NMR spectrum of aminosteroid derivative RM-133 dissolved in chloroform-d. All NMR signals were fully assigned using a combination of 2D NMR experiments
see Supporting Information).
(

7
4
R. Maltais et al. / Steroids 82 (2014) 68–76
0
0
including those we previously published for steroid derivatives
one of the two signals corresponding to the 5 -CH
whereas the two 2 -CH
and 46.1 ppm.
2
of proline,
1
0
[
22]. In H NMR (Fig. 2A), the most characteristic signals are 18-
CH (0.86 and 0.88 ppm), 19-CH (0.82 and 0.84 ppm) and 21-CH
2.57 and 2.58 ppm). The other signals from the steroid backbone
2
appear as four signals at 42.0, 42.5, 45.3
3
3
(
The left part of RM-133 is a side chain that contains the amino
acid proline and a quinoline moiety. In H NMR, the 2 -CH of pro-
line generated two very characteristic signals at 5.12 and
appears as a series of four multiplets be-
2 2
and 4 -CH are lo-
cated in a crowded area of the spectrum (1.9–2.4 ppm). In
1
00
all appear between 0.7 and 3.9 ppm, but they are not very charac-
teristic because of the overlapping of different signals. An alterna-
tive for the characterization of steroids is the use of 13C NMR
spectroscopy, which is a good source of information reported on
00
5.88 ppm. The 5 -CH
2
0
0
00
tween 3.8 and 4.2 ppm whereas those of 3 -CH
1
13
a larger scale (0–230 ppm) than H NMR (0–12 ppm). Thus, a com-
C
parison with the literature data allowed the rapid identification of
carbons 5–18, 20 and 21 (Fig. 3). It is noteworthy that the chemical
shift values of these carbons are not affected by the C2 and C3 sub-
NMR, all proline signals are duplicated and are generally of differ-
ent intensity. The more characteristic signals are those of the
0
0
amide carbonyl (1 -C) at 170.0 and 170.3 ppm, but the other car-
0
0
stituents. It is also important to mention that the 2b-R and 3
stereochemistries of RM-133 are deduced from the trans-diaxial
opening of the intermediate steroidal 2,3 -epoxide 2, and that this
conclusion has already been confirmed by X-ray analysis of the
7b-OH analog of 2 [8]. For the remaining carbons (1–4 and 19),
a-OH
bones are found at 57.4 and 59.0 ppm (2 -CH), 48.2 and
0
0
00
49.8 ppm (5 -CH
2
), 28.9 and 31.5 ppm (3 -CH
2
) and 22.4 and
0
0
a
25.3 ppm (4 -CH
line are located between 7.5 and 8.5 ppm (in H) and between
2
). As expected, the aromatic signals of the quino-
1
1
3
1
120.0 and 155.0 ppm (in C).
The presence of -proline and two amide bonds complexifies the
it was necessary to use 2D-NMR experiments, such as HSQC,
HMBC, COSY, TOCSY and NOESY [23], to identify these carbons.
These experiments are also helpful for the assignment of all car-
bons and protons of RM-133, which are reported in the experimen-
tal section. As an example, the HSQC experiment reported in Fig. 4
provides correlations between a carbon and geminal protons.
The middle part of RM-133 is constituted of a piperazine nu-
L
NMR signals of quinoline, which are duplicated. In fact, the pres-
ence of two conformers in proportions that range according to
the solvent used is clearly observed in Fig. 2A–D. The presence of
two conformers related to proline has been also observed by
NMR experiments in peptides derivatives [24–26]. Among the
1
00
large series of H NMR signals, we selected the CH-2 signals of
1
13
000
cleus. Both H and C NMR data of the four CH
2
are however
proline and the CH-4 signals of naphthyl group to illustrate the ef-
slightly interesting because they generally appear as broad signals
fect of solvent on both chemical shifts and conformer proportions
(Table 1). Thus, the 2 -CH signals ranged from 5.04 to 5.17 ppm for
1
13
1
00
in H and as a series of different intensity peaks in C. Thus, in H
0
NMR, the two 1 -CH
1
2
(part of an amine functionality) appear at
the minor conformer (I) and from 5.68 to 5.93 ppm for the major
0
00
.9–2.8 ppm whereas the two 2 -CH
appear at 3.1–3.8 ppm. In C NMR, the two 1 -CH
2
(part of amide functionality)
conformer (II). Using the intensity of these two 2 -CH signals, it
13
0
2
generated a ser-
is possible to estimate the proportions of both conformers to about
ies of weak signals located in the same area (47.7–48.4 ppm) than
3 3 3 3
46:54 in CDCl , 29:71 in CD OD and 25:75 in CD COCD or
Fig. 4. Partial representation of HSQC experiment (0–4.5 ppm for 1H NMR and 0–70 ppm for 13C NMR) of aminosteroid derivative RM-133 (in CDCl
) showing correlations
3
between a carbon and geminal protons. From the assignment of each carbon in the Y axis, it is possible to identify the corresponding geminal protons (X axis), which
sometimes generate duplicate or complex signals. The reverse strategy allows the identification of an unknown carbon signal from a known proton signal. See Fig. 3 for the
chemical structure and numbering of RM-133.

R. Maltais et al. / Steroids 82 (2014) 68–76
75
Table 1
Two characteristic 1H NMR signals showing the presence (d) and proportions (P) of two conformers of RM-133 in four different solvents.
Signal
CDCl
3
(chloroform-d)
CD
3
OD (methanol-d
4
)
CD
3
COCD
3
(acetone-d
6
)
3 3 6
CD SOCD (dimethyl-sulfoxide-d )
d in ppm (P in %)
d in ppm (P in %)
d in ppm (P in %)
d in ppm (P in %)
0
0
CH-2 (proline)-I
5.12 (45.7)
5.88 (54.3)
8.24 (44.8)
8.22 (55.2)
5.17 (29.0)
5.79 (71.0)
8.47 (28.6)
8.40 (71.4)
5.15 (25.4)
5.93 (74.6)
8.44 (24.8)
8.37 (75.2)
5.04 (25.4)
5.68 (74.6)
8.49 (25.9)
8.41 (74.1)
00
CH-2 (proline)-II
000
CH-4 (naphthyl)-I
000
CH-4 (naphthyl)-II
000
CD
3 3
SOCD . The 4 -CH signals ranged from 8.22 to 8.41 ppm for the
major conformer and from 8.24 to 8.49 ppm for the minor con-
000
former. A 45:55 conformer ratio and two 4 -CH signals shielded
were obtained in chloroform-d, whereas a 71:29 or a 75:25 ratio
and two signals slightly deshielded characterize the spectra ob-
tained in other solvents (methanol-d
4 6
, acetone-d and dimethyl-
sulfoxide-d ). It is interesting to mention that the ethylenic
6
proton (21-CH) located in the right side of RM-133 is also influ-
enced by the two conformers although the proline–quinoline side
chain is located far from 21-CH in the left side of the molecule.
We also observed that the methyl groups (18 and 19) are both
shielded in dimethylsulfoxide-d
more influenced by the two conformers than the 18-CH
6
. As expected, the 19-CH
3
is much
, but this
3
effect on the chemical shift is less important in chloroform-d. Fi-
1
nally, the H NMR spectra of RM-133 in four different solvents
(Fig. 2) clearly demonstrate how complex the NMR data of a mol-
ecule such as RM-133 can be.
Fig. 5. Efficacy of RM-133 to reduce the growth of tumors (xenograft model with
human leukemia (HL-60) cells) in nude mice. The aminosteroid RM-133 (60 mg/kg)
3.3. Anticancer activity of RM-133
in 0.2 mL of propylene glycol:ethanol (92:8) was injected subcutaneously once per
⁄
2
days. P < 0.01 (RM-133-treated group vs. control (Intact) group).
The in vitro anticancer potential of RM-133 has been reported
on a panel of cancer cell lines demonstrating interesting activity
IC50) at a low micromolar level, notably on leukemia (HL-60;
propylene glycol:ethanol may allow us to increase the bioavailabil-
ity of the RM-133 thereby increasing its effectiveness.
(
IC50 = 0.1
IC50 = 0.1
l
l
M), breast (MCF-7; IC50 = 0.1
M), prostate (LNCaP; IC50 = 2.0
l
M), breast (T-47D;
4
. Conclusions
l
M) cancer cells [14].
Furthermore, a preliminary investigation of the mechanisms of ac-
tion of this family of aminosteroid derivatives has also been initi-
ated, pointing toward a pro-apoptotic inducing action that works
through the extrinsic pathway [4]. Since we were able to generate
multi-gram quantities of RM-133, we were now interested in eval-
uating the potency of RM-133 to counter tumor growth in a xeno-
graft model in nude mice. We chose human leukemia (HL-60) cells
to proceed with a xenograft on nude mice, considering that our
preliminary data on the mechanisms of action had been performed
on HL-60 cells [4]. For this in vivo assay, Balb/c nude mice were
inoculated with HL-60 cells (2 million) in each flank. Only mice
that developed well-established tumors were selected to continue
the study. Mice were then randomized by tumor size and divided
into two groups: the first treated with RM-133 (60 mg/kg) in a sus-
pension of 0.1 mL propylene glycol–ethanol (92:8) once every
The aminosteroid derivative RM-133 was developed from struc-
ture–activity relationship studies. This new steroid derivative
showed promising anticancer activity on a panel of cancer cell
lines, for which in vitro results justified the first preclinical studies.
In order to proceed with the in vivo evaluation of RM-133 on a
xenograft model, we optimized its chemical synthesis to obtain
multi-gram quantities. Using a convergent synthesis, this ADT
derivative was thus obtained in six chemical steps and only three
chromatographic purifications for an overall yield of 14%. The syn-
thesis was realized on multi-gram scale and performed in 2 weeks
by a single chemist. The NMR analysis revealed that RM-133 exists
as a mixture of two conformers of which proportions are affected
by the solvents. The presence of two amide bonds linking a piper-
azine, a proline and a quinoline explain these conformers. This
interesting observation by NMR about two conformers provided
additional information on the molecular structure of RM-133,
which sheds new light on SAR parameters that could influence bio-
logical activity. The evaluation of RM-133 activity on a HL-60 xeno-
graft showed an interesting result with blocking of tumor
progression by 57% obtained under non-optimized conditions. This
work, including the development of a viable chemical route to RM-
2
days, whereas the second was treated with the vehicle only.
The surface of the tumor in the intact group increased steadily dur-
ing the first 4 weeks. Thereafter, the tumor surface increased rap-
2
idly during the following week to an average size of 160 mm , or
1
6 times the original size. Meanwhile, the surface of tumors trea-
ted with RM-133 increased much slower than for the intact group,
as tumor size at the end of treatment was more than twice as large.
Indeed, the group treated with RM-133 showed an interesting
in vivo activity by blocking the progression of HL-60 tumors by
1
33, a better characterization of this aminosteroid derivative by
NMR and a preliminary in vivo proof-of-concept of anticancer
activity (xenograft model), will help progress toward clinical
studies.
5
7% (P < 0.01) after 39 days of treatment (Fig. 5). It is also impor-
tant to mention that no apparent toxicity was observed during
the entire treatment period, including no body weight loss. These
results, although very interesting, demonstrate a partial blocking
of tumor growth. It is reasonable to believe that the use of a higher
dose would allow us to obtain a larger or even total block of tumor
growth. In addition, the use of another vehicle than the mixture of
Acknowledgments
We thank the Canadian Institutes of Health Research (CIHR) and
the Québec Breast Cancer Research Foundation (QBCRF) for their

7
6
R. Maltais et al. / Steroids 82 (2014) 68–76
financial support, and Ms. Micheline Harvey for careful reading of
the manuscript.
[12] Djerassi C. Chemical birth of the pill. Am J Obst Gynecol 2006;194:291–8.
[
13] Talbot A, Maltais R, Poirier D. New diethylsilylacetylenic linker for parallel
solid-phase synthesis of libraries of hydroxyl acetylenic steroid derivatives
with improved metabolic stability. ACS Comb Sci 2012;14:347–51.
[
[
[
14] Ayan D. Activité biologique in vitro et in vivo de dérivés stéroïdiens pour le
traitement du cancer. Université Laval [M.Sc. thesis, chapter 6], p. 142–166.
15] Kramer CY. Extension of multiple range tests to group means with unequal
numbers of replications. Biometrics 1956;12:307–10.
16] Anderson A, Boyd AC, Byford A, Campbell AC, Gemmell DK, Hamilton NM, et al.
Anesthetic activity of novel water-soluble 2 beta-morpholinyl steroids and
their modulatory effects at GABAA receptors. J Med Chem 1997;40:1668–81.
17] Hudlicky M. Fluorination with diethylaminosulfur trifluoride and related
aminofluorosulfuranes. Org React 1988;35:513–641.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.steroids.2014.
0
1.008.
[
[
References
18] Bird TGC, Fredericks PM, Jones ERH, Meakins GD. Convenient general
preparation of oxygenated monofluoro- and gem-difluoro-5
using diethylaminosulphur trifluoride. J Chem Soc 1979:65–6.
a-androstanes
[
[
1] Atkin JH, Gershell LJ. Selective anticancer drugs. Nat Rev Cancer 2002;2:645–6.
2] Raguz SY, Yagüe E. Resistance to chemotherapy: new treatments and novel
insights into an old problem. Br J Cancer 2008;99:387–91.
[
[
19] Blunt JW, Stothers JB. 13C NMR spectra of steroids – a survey and commentary.
Org Magn Reson 1977;9:439–64.
[
[
3] Burgess DJ. Anticancer drugs: selective oxycution? Nat Rev Drug Discov
20] Hunter AC, Khuenl-Brady H, Barrett P, Dodd HT, Dedi C. Transformation of
2
011;10:658–9.
4] Jegham H, Roy J, Maltais R, Desnoyers S, Poirier D. A novel aminosteroid of the
-androstane-3 ,17b-diol family induces cell cycle arrest and apoptosis in
human promyelocytic leukemia HL-60 cells. Invest New Drugs
012;30:176–85.
some
3a-substituted steroids by Aspergillus tamarii KITA reveals
stereochemical restriction of steroid binding orientation in the minor
hydroxylation pathway. J Steroid Biochem Mol Biol 2010;118:171–6.
21] Hunter AC, Khuenl-Brady H, Barrett P, Dodd HT, Dedi C. Transformation of a
series of saturated isomeric steroidal diols by Aspergillus tamarii KITA reveals a
precise stereochemical requirement for entrance into the lactonization
pathway. J Steroid Biochem Mol Biol 2010;122:352–8.
5a
a
[
[
2
[
[
5] Jegham H, Maltais R, Roy J, Doillon C, Poirier D. Biological evaluation of a new
family of aminosteroids that display a selective toxicity for various malignant
cell lines. Anti-Cancer Drugs 2012;23:803–14.
6] He Q, Jiang D. A novel aminosteroid is active for proliferation inhibition and
differentiation induction of human acute myeloid leukemia HL-60 cells. Leuk
Res 1999;23:369–72.
22] Tchédam Ngatcha B, Trottier MC, Poirier D. 13C Nuclear magnetic resonance
spectroscopy data of
derivatives substituted at position 3beta or/and 3alpha. Curr Top Steroid Res
011;8:35–46.
a variety of androsterone and epi-androsterone
2
[
[
7] He Q, Na X. The effects and mechanisms of a novel 2-aminosteroid on murine
WEHI-3B leukemia cells in vitro and in vivo. Leuk Res 2001;25:455–61.
[
[
23] Claridge TDW. High-resolution NMR techniques in organic chemistry. In:
Baldwin JE, Williams FRS, Williams RM, editors. Tetrahedron organic
chemistry series, vol. 19. Amsterdam: Pergamon; 1999.
8] Roy J, DeRoy P, Poirier D. 2b-(N-substituted piperazino)-5
,17b-diols: parallel solid-phase synthesis and antiproliferative activity on
human leukemia HL-60 cells. J Comb Chem 2007;9:347–58.
9] Thibeault D, Roy J, DeRoy P, Poirier D. Chemical synthesis of 2b-amino-5
androstane-3 ,17b-diol N-derivatives and their antiproliferative effect on HL-
0 human leukemia cells. Bioorg Med Chem 2008;16:5062–77.
10] Roy J, Maltais R, Jegham H, Poirier D. Libraries of 2b-(N-substituted
piperazino)-5 -androstane-3 ,17b-diols: chemical synthesis and cytotoxic
a-androstane-
3a
24] Sikoroska E, Sobolewski D, Kwiatkowska A. Conformational preferences of
proline derivatives incorporated into vasopressin analogues: NMR and
molecular modelling studies. Chem Biol Drug Des 2012;79:535–47.
[
a-
a
[
[
25] Abraham RJ, Hudson BD. 1H NMR spectrum and conformation of the proline
ring in propionylalanylprolineethylamide. Magn Reson Chem 1986;24:812–5.
26] Haasnoot CAG, DeLeeuw FAAM, DeLeeuw HPM, Altona C. Relationship
between proton–proton nmr coupling constants and substituent
electronegativities. III. Conformational analysis of proline rings in solution
using a generalized Karplus equation. Biopolymers 1981;20:1211–45.
6
[
[
a
a
effects on human leukemia HL-60 cells and on normal lymphocytes. Mol
Divers 2011;15:317–39.
11] Jegham H, Maltais R, Dufour P, Roy J, Poirier D. Solid-phase chemical synthesis
and in vitro biological evaluation of novel 2b-piperazino-(20R)-5
a-pregnane-
3a
,20-diol N-derivatives as anti-leukemic agents. Steroids 2012;77:1403–18.