Synthesis of [19-2H3]-analogs of dehydroepiandrosterone and pregnenolone and their sulfates

Article abstract of DOI:10.1016/j.steroids.2003.11.002

Deuterated analogs of pregnenolone and pregnenolone sulfate with three atoms of deuterium in position 19 were prepared. The synthetic approach was developed on derivatives of dehydroepiandrosterone, where initial intermediates were well characterized, and

Full text of DOI:10.1016/j.steroids.2003.11.002

Steroids 69 (2004) 161–171
Synthesis of [19-2H3]-analogs of dehydroepiandrosterone and
pregnenolone and their sulfates
a
a,∗
a
b
ˇ
´
ˇ ´
ˇ´
Ivan Cerný , Vladimır Pouzar , Miloš Budešınský , Marie Bicıková ,
Martin Hillb, Richard Hamplb
a
Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic, 166 10 Prague 6, Czech Republic
b
Institute of Endocrinology, Prague, Czech Republic
Received 3 September 2003; received in revised form 7 November 2003; accepted 12 November 2003
Abstract
Deuterated analogs of pregnenolone and pregnenolone sulfate with three atoms of deuterium in position 19 were prepared. The syn-
thetic approach was developed on derivatives of dehydroepiandrosterone, where initial intermediates were well characterized, and then
applied to the pregnenolone series. Starting 19-hydroxy compounds were transformed into 3␣,5-cycloderivatives to simplify the Jones
oxidation into the corresponding 19-oic acids. After oxidation, rearrangement to 3-hydroxy-5-enes, and suitable protection, two deuterium
atoms were introduced by lithium aluminum deuteride reduction. Mesylate exchange by iodide in the presence of zinc and deuterium
oxide added third deuterium atom. Deprotection gave title analogs with about 93–95% content of d3-derivative, the rest was mainly not
fully deuterated d2-analogue as followed from the mass spectra analysis. Thus, 3␤-hydroxy[19-2H3]androst-5-en-17-one was prepared
in 14 steps from 19-hydroxy-17-oxoandrost-5-en-3␤-yl acetate in 8.9% yield, the analogous sequence in the pregnenolone series gave
3␤-hydroxy[19-2H3]pregn-5-en-20-one in 7.3% yield. Corresponding sulfates were prepared via pyridinium salts in 53 and 57% yields,
respectively. Fully assigned NMR data of selected pregnenolone derivatives were given.
© 2004 Elsevier Inc. All rights reserved.
Keywords: Steroid; Synthesis; Deuterium; NMR spectra
1. Introduction
requirement when using deuterium labeling is the pres-
ence of at least two deuterium atoms substituting protium.
This allows for sufficient sensitivity, which is influenced
by the interference of the first isotopic satellite caused
by the natural abundance of 13C isotope. Moreover, it
is necessary to use such positions on the steroid skele-
ton where deuterium is sufficiently stable for protium
exchange and for possible metabolic transformations of
the steroid skeleton under study. Taking this together, la-
beling with three deuterium atoms at the angular methyl
in position 19 fulfills all prerequisites for successful
tracing.
With respect to the development of a synthetic ap-
proach for the above-mentioned deuterated derivatives
of DHEA and PREG (Scheme 1), substantial back-
ground information exists in the literature. The function-
alization of position 19 was feasible through lead(IV)
acetate cyclization of the corresponding 6␤-hydroxy
derivatives [7–9]. Transformations into 19-oic acid
were performed mostly on 3-oxo-4-ene derivatives [8],
and the use of 6␤-methoxy-3␣,5-cycloderivative was
previously described in patent literature [10]. Introduction
Recent interest in neuroactive steroids stimulated the need
for compounds that can be traced in biological tissues to-
gether with their metabolites. Our study focused on de-
hydroepiandrosterone (DHEA), pregnenolone (PREG), and
their sulfates. These derivatives are included among neuros-
teroids, i.e., they are both synthesized and accumulated in
the nervous system to levels, at least in part, independent of
peripheral steroidogenesis. Neurosteroids interact with vari-
ous brain cell receptors, mainly through ␥-aminobutyric acid
(GABA) receptors or N-methyl-d-aspartate (NMDA) gluta-
matergic receptors. They have been demonstrated to exhibit
a broad spectrum of biological actions, influencing memory
and mental activity (e.g., learning, perception, cognition, and
spatial orientation) [1–5].
Among the analytical methods for tracing neurosteroids,
a combination of gas chromatography–mass spectroscopy
combination is the frequently used one [6]. The principal
∗
Corresponding author. Tel.: +420-220-183-111;
fax: +420-220-183-559.
E-mail address: pouzar@uochb.cas.cz (V. Pouzar).
0039-128X/$ – see front matter © 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.steroids.2003.11.002

ˇ
I. Cerny´ et al. / Steroids 69 (2004) 161–171
162
complete structural assignment of signals (Tables 1 and 2).
Mass spectra were recorded by means of a VG Analytical
ZAB-EQ spectrometer.
Thin-layer chromatography (TLC) was performed on sil-
ica gel G (ICN Biochemicals, detection by spraying with
concentrated sulfuric acid followed by heating). For col-
umn chromatography, Silica gel 60 (Merck, 63–100 ␮m)
or Aluminum oxide 90 neutral (Merck, 63–200 ␮m, ac-
tivity III) were used. Prior to evaporation on a rotary
evaporator in vacuo (0.25 kPa, bath temperature 40 ◦C),
solutions in organic solvents were dried over anhydrous
MgSO4.
19-Hydroxy-3␤-methoxy-3␣,5-cyclo-5␣-androstan-17-one
(1) was prepared both according to Ref. [13] and by a
procedure described below for PREG derivative 18 (yield
from 19-hydroxy-17-oxoandrost-5-en-3␤-yl acetate was
Scheme 1.
of two deuterium atoms using lithium aluminum deu-
teride reduction of methyl ester is a standard method;
in the androstane series, it was performed on methyl
17,17-ethylenedioxy-3␤-methoxyandrost-5-en-19-oate [11].
For the third deuterium atom, a general method using an
exchange of the mesyloxy group could be applied [12].
However, some of the existing methods are specific for a
particular type of steroid (mostly for 3-oxo-4-enes of the
androstane series in connection with aromatization studies),
and especially in the PREG series, even initial intermediates
are not known or sufficiently documented. We, therefore,
decided to develop our own approach using the androstane
series for preliminary experiments and applied the sequence
to the PREG series.
36%), m.p. 102–103 ◦C, [α] + 128 (c 1.2, CHCl3), liter-
d
ature [13] gave m.p. 104–107 ◦C, [α] + 145.2 (CHCl3);
d
19-hydroxy-20-oxopregn-5-en-3␤-yl acetate (13) was
prepared according to Ref. [9], m.p. 171–173 ◦C (ace-
tone/hexane), [α] + 22 (ca. 0.6, CHCl3), 1H and 13C
d
NMR, see Tables 1 and 2, literature gives m.p. 165–176 ◦C,
[α] + 19 [7] and m.p. 168–171 ◦C (diethyl ether/pentane)
d
[9]; pregnenolone and pregnenolone acetate (PREG-Ac)
were obtained from Steraloids. Jones reagent was pre-
pared by dissolving CrO3 (13.35 g) in concentrated H2SO4
(11.5 ml) and water (20 ml) and subsequently diluting to
50 ml with water. Lithium aluminum deuteride (Aldrich)
had 98% LiAlD4. For deuterium exchange experiments
(9 and 28), all solid components were predried in vacuo:
powdered mesylate at 50 ◦C for 5 h, zinc and sodium io-
dide at 105 ◦C for 10 h. 1,2-Dimethoxyethane was left
for several days over 4A molecular sieves, and it had
a residual water content of 30 ␮g/ml. Deuterium oxide
(CDN isotopes, Quebec, Canada) had a 99.9% content of
D2O.
2. Experimental
2.1. General
Melting points were determined on a Boetius micro-
melting point apparatus (Germany). Optical rotations were
measured at 25 ◦C on a Perkin-Elmer 141 MC polarime-
ter, and [α]D values are given in 10−1 deg cm2 g−1. In-
frared spectra (wavenumbers in cm−1) were recorded on a
2.2. Derivatives of DHEA
2.2.1. Methyl 3β-hydroxy-17-oxoandrost-5-en-19-oate (4)
Freshly prepared Jones reagent (2.67 M, 7 ml) was added
dropwise to a stirred solution of cycloderivative 1 (1.4 g,
4.40 mmol) in acetone (90 ml) under argon and in an ice
bath. The reaction mixture was stirred and cooled for 3.5 h
(TLC: chloroform/acetone (10:1)). 2-Propanol (1 ml) was
then added, the mixture was stirred while cooling for 5 min,
and then poured into saturated aqueous KHCO3 (100 ml).
Part of the acetone was removed on a rotary evaporator, ad-
ditional KHCO3 (100 ml) was added, and the product was
extracted with ethyl acetate (2× 150 ml). The organic lay-
ers were collected, dried, and the solvent was evaporated.
The crude acid 2 (1.2 g) was dissolved in acetone (25 ml),
35% perchloric acid (2.5 ml) was added, and the mixture
was stirred at room temperature for 1 h. After pouring into
ice-water (100 ml), the crystalline product was filtered off,
washed with water, and dried under air. Crude 3 (1.0 g) was
1
Bruker IFS 88 spectrometer. H NMR spectra were taken
on Varian UNITY-200 and Bruker AVANCE-400 instru-
ments (200 and 400 MHz, FT mode) at 23 ◦C in CDCl3
and referenced to TMS as the internal standard. Chemical
shifts are given in ppm (δ-scale); coupling constants (J)
and width of multiplets (W) are given in Hz. NMR spec-
tra for selected pregnane derivatives (13, 15, 16, 18, 23,
29–32, PREG, and PREG-Ac) were measured on Varian
UNITY-500 and Bruker AVANCE-500 instruments (1H
at 500 MHz; 13C at 125.7 MHz) under the above condi-
tions, except for 32 that was measured in D2O. For 13C,
secondary referencing was performed using the solvent
signal at position δ(CDCl3) = 77.0. In addition to 1D
proton and carbon NMR spectra, homonuclear 2D-spectra
(H,H-PFG-COSY, ROESY and J-resolved) together with
heteronuclear 2D-spectra (H,C-PFG-HSQC) were used for

ˇ
I. Cerny´ et al. / Steroids 69 (2004) 161–171
163
Table 1
1H NMR chemical shifts in CDCl3 (except for 32 which was measured in D2O at 50 ◦C)
Proton
13 (a)
15 (b)
16 (c)
18 (d)
23 (e)
29
30
1.10
1.85
1.85
1.50
3.53
2.31
2.23
5.36
1.48
2.00
1.48
0.99
∼1.62
∼1.62
1.46
2.05
1.15
31 (f)
32
PREG
PREG-Ac (g)
1␣ (ax)
1␤ (eq)
2␣ (eq)
2␤ (ax)
3
4␣ (eq)
4␤ (ax)
6
7␣ (ax)
7␤ (eq)
8 (ax)
9 (ax)
11␣ (eq)
11␤ (ax)
12␣ (ax)
12␤ (eq)
14 (ax)
15␣
1.16
1.97
1.88
1.50
4.65
2.43
2.27
5.78
1.58
2.06
1.88
0.98
∼1.66
∼1.66
1.45
2.09
1.07
1.66
1.24
2.18
∼1.65
2.53
0.684
3.87
3.61
2.125
1.11
2.08
1.87
1.41
3.56
2.37
2.26
5.60
1.55
2.04
1.74
1.00
1.69
1.51
1.43
2.08
1.08
1.68
1.24
2.18
1.66
2.52
0.651
4.47
3.96
2.118
1.11
1.95
1.86
1.40
3.58
2.39
2.20
5.75
1.57
2.06
1.90
0.96
∼1.67
∼1.67
1.43
2.09
1.06
0.71
1.34
1.76
1.57
1.13
0.64
0.48
2.80
1.07
1.97
2.11
0.87
1.55
1.76
1.43
2.08
1.12
1.67
1.24
2.21
1.64
2.54
0.723
3.53
1.03
2.61
1.95
1.51
3.55
2.48
1.79
5.66
1.59
2.13
1.79
1.10
1.78
1.09
1.43
2.04
1.11
1.69
1.25
2.18
1.66
2.51
0.567
–
1.10
1.13
1.87
2.18
1.70
4.36
2.64
2.42
5.39
1.55
1.99
1.48
0.99
1.61
1.45
1.46
2.04
1.15
1.67
1.23
2.17
1.66
2.53
0.625
–
1.15
1.96
2.04
1.70
4.20
2.53
2.42
5.52
1.61
2.03
1.51
1.04
1.68
1.52
1.52
2.05
1.27
1.73
1.26
2.04
1.73
2.74
0.63
–
1.10
1.86
1.85
1.50
3.53
2.31
2.24
5.35
1.59
2.00
1.48
0.99
1.63
1.48
1.46
2.05
1.15
1.68
1.24
2.17
1.65
2.53
0.634
1.011
1.16
1.87
1.87
1.60
4.61
2.34
2.31
5.38
1.58
2.00
1.48
1.01
1.62
1.48
1.47
2.05
1.18
1.68
1.23
2.18
1.65
2.54
0.632
1.022
1.86
1.86
1.50
3.53
2.29
2.24
5.36
1.58
2.00
1.48
0.99
1.63
1.48
1.46
2.05
1.16
1.68
1.24
2.18
∼1.65
2.53
0.634
–
∼1.65
∼1.65
15␤
16␣
16␤
17
18
19
19ꢀ
1.24
1.24
2.18
2.18
∼1.65
∼1.65
2.51
2.53
0.688
3.84
3.60
0.634
0.976
21
2.118
2.126
2.103
2.127
2.127
2.125
2.21
2.127
2.127
Other protons—(a) OAc: 2.036 s; (b) OAc: 2.027 s; (c) OH: 1.05 b; (d) OH: 4.40 dd; (e) OMe: 3.702 s; (f) C5H5N: 8.98 m (H-2,6), 8.00 m (H-3,5),
8.49 m (H-4); (g) OAc: 2.037 s.
Table 2
13C NMR chemical shifts in CDCl3 (except for 32 which was measured in D2O at 50 ◦C)
Carbon
13 (a)
15 (b)
16
18 (c)
23 (d)
29
30
31 (e)
32
36.89
PREG
PREG-Ac (f)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
33.14
28.04
73.28
38.13
134.56
127.95
31.08
33.25
50.14
41.57
21.67
39.03
44.15
57.68
24.27
22.79
63.60
13.55
62.69
209.50
31.49
33.77
31.77
71.37
42.36
135.64
125.51
31.22
32.92
50.09
39.64
21.69
38.99
44.01
57.46
24.40
22.84
63.60
13.28
64.53
209.28
31.43
33.40
29.80
24.08
23.30
12.53
35.58
82.01
32.94
32.06
47.39
48.44
22.54
39.50
44.76
57.44
24.24
22.81
63.78
13.86
66.50
209.52
31.46
33.83
33.04
70.95
44.17
134.82
124.39
30.85
32.03
48.64
50.73
23.32
38.64
43.92
56.54
24.26
22.81
63.54
13.16
174.10
209.28
31.46
37.17
37.20
37.07
28.83
78.80
39.17
140.14
122.04
31.77
31.82
49.83
36.29
21.03
38.79
43.98
56.84
24.46
22.79
63.70
13.20
18.42
209.59
31.54
37.28
31.63
71.70
42.27
140.79
121.37
31.78
31.87
50.02
36.53
21.09
38.86
43.99
56.93
24.48
22.85
63.72
13.21
19.37
209.46
31.49
37.01
27.73
73.83
38.07
139.68
122.31
31.76
31.83
49.92
36.61
21.03
38.80
43.96
56.85
24.47
22.85
63.68
13.20
19.28
209.42
31.50
31.09
71.30
42.24
135.57
127.04
31.90
33.31
50.23
41.50
21.72
39.09
44.19
57.78
24.28
22.79
63.63
13.57
62.72
209.56
31.47
31.60
71.70
42.24
140.74
121.38
31.75
31.86
49.93
36.30
21.08
38.63
43.99
56.90
24.47
22.80
63.70
13.21
18.51
209.58
31.54
31.61
71.70
42.24
140.73
121.40
31.76
31.86
49.94
36.37
21.08
38.83
44.00
56.90
24.48
22.80
63.70
13.22
18.80
209.58
31.54
28.61
80.73
38.86
141.60
122.17
31.65
31.86
49.92
36.68
21.01
38.40
44.65
56.60
24.30
22.87
63.92
12.82
17.19
217.93
31.35
g
h
g
g
(a) OAc: 170.50, 21.36; (b) OAc: 170.63, 21.03; (c) OMe: 56.56; (d) OMe: 51.80; (e) C6D5NH: 142.28 (C-2,6), 127.12 (C-3,5), 145.69 (C-4); (f) OAc:
170.49, 21.38; (g) heptet (1J(C, D) = 18.7 Hz); (h) pentet (1J(C, D) = 19.3 Hz).

ˇ
I. Cerny´ et al. / Steroids 69 (2004) 161–171
164
dissolved in methanol (30 ml), cooled with an ice bath, and
methylated with ethereal diazomethane solution (to yellow
color). After 5 min (TLC: chloroform/acetone (10:1)), sev-
eral drops of acetic acid were added, and the solution was
evaporated to dryness (1.2 g). Column chromatography on
silica gel (150 ml) in a mixture of chloroform/acetone (20:1)
3.33 mmol) under argon for 4 h. The reaction mixture was
then chilled in an ice bath, the remaining deuteride was
destroyed with saturated, aqueous Na2SO4, and the mixture
was diluted with ether (15 ml). The solids were filtered off
on a celite column (10 ml), which was then washed with
ether (50 ml). The resulting solution was washed sequen-
tially with cold 5% aqueous citric acid, water, saturated
aqueous KHCO3 (2×), and water. After drying, the sol-
vents were evaporated, leaving 460 mg (98%) of hydroxy
gave 999 mg of ester 4 (68%), m.p. 188–190 ◦C, [α] − 61
d
(c 0.6, CHCl3). Literature [14] gave m.p. 188–190 ◦C. IR
(CHCl3), ν (cm−1): 3607, 3480 (OH); 1733, 1723 (C O,
=
ester, ketone); 1677 (C C); 1438 (CH3, ester); 1206, 1167
derivative 7, m.p. 140–141 ◦C (methanol), [α] − 64 (c 0.2,
=
d
1
(C–O). H NMR (200 MHz), δ (ppm): 5.69 (1H, m, W =
CHCl3). IR (CHCl3), ν (cm−1): 3624, 3577 (OH); 2217,
2
=
12, H-6); 3.73 (3H, s, COOCH3); 3.57 (1H, m, W = 30,
H-3␣); 0.82 (3H, s, 3 × H-18). Analysis calculated for
C20H28O4 (332.4): C, 72.26; H, 8.49. Found: C, 72.37; H,
8.53.
2116 (C H2); 1667 (C C); 1149, 1104, 1042 (COCOC,
methoxymethoxy); 1104, 959, 950 (ring, dioxolane). 1H
NMR (200 MHz), δ (ppm): 5.76 (1H, m, W = 12, H-6);
4.69 (2H, s, O–CH2–O); 3.87 (4H, m, OCH2CH2O); 3.48
(1H, m, W = 32, H-3␣); 3.37 (3H, s, CH3–O–CH2); 0.92
(3H, s, 3 × H-18). Analysis calculated for C23H342H2O5
2.2.2. Methyl 17,17-ethylenedioxy-3β-(methoxymethoxy)
androst-5-en-19-oate (6)
2
(394.6): C, 70.02; H, 8.69; H, 1.02. Found: C, 69.94; H,
Ethylene glycol (1.5 ml, 27 mmol), triethyl orthoformate
(4.5 ml, 27 mmol), and 4-toluenesulfonic acid monohydrate
(10 mg) were added to a suspension of ester 4 (900 mg,
2.71 mmol) in benzene (20 ml). The mixture was stirred at
room temperature for 5 h (TLC: benzene/ether (1:1)), and
during this time, the suspension dissolved. The reaction mix-
ture was poured into ice cooled, saturated aqueous KHCO3
(100 ml) and extracted with ethyl acetate (100 ml). The ex-
tract was washed successively with KHCO3 (2×), and with
water, dried, and the solvent was evaporated. Crude ketal 5
(960 mg) was dissolved in benzene (20 ml), cooled in an ice
bath, and N,N-diisopropylethylamine (5 ml, 28.7 mmol) and
bromomethyl methyl ether (0.6 ml, 7.4 mmol) were added.
The mixture was stirred at room temperature for 2 h (TLC:
benzene/acetone (10:1)). Then, the product was partitioned
between ethyl acetate (100 ml) and cold 5% aqueous cit-
ric acid (100 ml). The organic layer was washed sequen-
tially with 5% aqueous citric acid, water, saturated aqueous
KHCO3 (2×), and water, dried, and the solvent was evap-
orated. The residue (800 mg) in minimum benzene was ap-
plied onto a short column of alumina (30 ml) in benzene, and
the product was eluted with a mixture of benzene/acetone
(20:1). The yield of protected ester 6 was 590 mg (52%),
9.44 (distorted by 2H).
2.2.4. 17,17-Ethylenedioxy-3β-(methoxymethoxy)[19-2H3]
androst-5-ene (9)
Hydroxy derivative 7 (450 mg, 1.14 mmol) was dissolved
in pyridine (5 ml), chilled in an ice bath, and methanesul-
fonyl chloride (0.4 ml, 5.17 mmol) was added. The reaction
mixture was left aside in an ice bath for 1 h. Then it was
poured onto ice, and the solid product formed was filtered
off. The solids were dissolved in ether (50 ml), and the so-
lution was washed sequentially with ice cold, 5% aqueous
citric acid (2×), water, saturated aqueous KHCO3 (2×), and
water. After drying, the solvent was evaporated, the residue
was coevaporated with benzene (3× 10 ml), and thoroughly
dried in vacuo. Crude mesylate 8 (430 mg) was refluxed
under an argon atmosphere in 1,2-dimethoxyethane (4 ml)
with sodium iodide (430 mg, 2.87 mmol), zinc (430 mg,
6.58 mmol), and deuterium oxide (0.4 ml, 22.11 mmol) for
6 h. The reaction mixture was cooled, diluted with ether,
and filtered through a celite column (10 ml), which was
washed sequentially with ether (ca. 40 ml). The ethereal
solution was washed with water, ice cold 5% aqueous citric
acid (2×), water, saturated aqueous KHCO3, 5% aqueous
sodium thiosulfate pentahydrate, and water. After drying,
the solvent was evaporated. The residue was dissolved in
a minimum of benzene, introduced onto a short column
of aluminum oxide (15 ml), and the product was eluted
with a mixture of benzene/ether (10:1). The yield of pro-
tected d3-DHEA (9) was 340 mg (78%), m.p. 90–92 ◦C
m.p. 134–136 ◦C (methanol), [α] −118 (c 0.5, CHCl3). IR
d
(CHCl3), ν (cm−1): 1721 (C O, ester); 1675 (C C); 1437
(CH3, ester); 1169, 1031 (C–O, ester); 1150, 1106, 1041
(COCOC, methoxymethoxy); 1106, 959, 952 (ring, diox-
olane). 1H NMR (200 MHz), δ (ppm): 5.67 (1H, bd, J = 5.8,
H-6); 4.67 (2H, s, O–CH2–O); 3.88 (4H, m, OCH2CH2O);
3.71 (3H, s, COOCH3); 3.45 (1H, m, W = 30, H-3␣); 3.36
(3H, s, CH3–O–CH2); 0.79 (3H, s, 3 × H-18). Analysis cal-
culated for C24H36O6 (420.6): C, 68.55; H, 8.63. Found: C,
68.67; H, 8.63.
=
=
(methanol), [α] − 67 (c 0.2, CHCl3). IR (CHCl3), ν
d
2
(cm−1): 2225 (C H3); 1668 (C C); 1148, 1105, 1042 (CO-
=
COC, methoxymethoxy); 1105, 959, 951 (ring, dioxolane).
1H NMR (200 MHz), δ (ppm): 5.36 (1H, m, W = 12,
H-6); 4.69 (2H, s, O–CH2–O); 3.86 (4H, m, W = 30,
OCH2CH2O); 3.42 (1H, m, W = 31, H-3␣); 3.37 (3H, s,
CH3–O–CH2); 0.86 (3H, s, 3 × H-18). Analysis calculated
for C23H332H3O4 (379.6): C, 72.78; H, 8.76; 2H, 1.52.
Found: C, 72.72; H, 9.74 (distorted by 2H).
2.2.3. 17,17-Ethylenedioxy-3β-(methoxymethoxy)[19-2H2]
androst-5-en-19-ol (7)
Protected ester 6 (500 mg, 1.19 mmol) in THF (15 ml)
was refluxed with lithium aluminum deuteride (140 mg,

ˇ
I. Cerny´ et al. / Steroids 69 (2004) 161–171
165
2.2.5. 3β-Hydroxy[19-2H3]androst-5-en-17-one (10)
Protected d3-DHEA (9) (300 mg, 0.79 mmol) was stirred
in benzene (5 ml) and a methanol (5 ml) solution with 35%
aqueous HClO4 at 50 ◦C for 4 h. The reaction mixture was
cooled, poured into cold, saturated aqueous NaHCO3, and
extracted with ethyl acetate (20 ml). The extract was washed
with saturated aqueous NaHCO3 and water, dried, and the
solvents were evaporated. The residue was chromatographed
on silica gel (20 ml) in a mixture of benzene/acetone
(50:1). The yield of d3-DHEA (10) was 210 mg (91%),
(5 ml) and adding chloroform (7.5 ml). The yield of sodium
salt 12 was 210 mg (68%), m.p. 192–194 ◦C, [α] + 18 (c
d
0.2, methanol). Ref. [16] gave m.p. 192–193 ◦C and Ref.
[17] gave m.p. 185–187 ◦C, [α] + 10.7 (c 4, methanol) for
d
non-deuterated analog. IR (KBr), λ (cm−1): 2220 (C2H3);
=
1733 (C O); 1289–1218, 1065, 986, 969 (sulfate). Anal-
ysis calculated for C19H242H3NaO5S · H2O (411.5): C,
55.46; H, 6.37; 2H, 1.47; S, 7.79. Found: C, 55.16; H, 7.31
(distorted by 2H); S, 7.75.
m.p. 149–151 ◦C (methanol), [α] +12 (c 0.2, ethanol). Ref.
2.3. Derivatives of PREG
d
[15] gives m.p. 142–143 and 152–153 ◦C, [α] + 10.9 (c
d
0.4, ethanol) for the non-deuterated analog. IR (CHCl3), ν
2.3.1. 20-Oxopregn-5-ene-3β,19-diyl diacetate (14)
2
(cm−1): 3608, 3458 (OH); 2225 (C H3); 1732 (C O); 1668
=
Acetate 13 (6 g, 16.0 mmol) was dissolved in pyri-
dine (50 ml), the solution was cooled in an ice bath, and
4-dimethylaminopyridine (650 mg, 5.2 mmol) and acetic an-
hydride (7 ml, 74.2 mmol) were added. The reaction mixture
was left at room temperature for 3 h (TLC: benzene/acetone
(10:1)). Then, the reaction mixture was poured into an ice
and water mixture (500 ml), and the crystals were filtered
off and washed with water. The crude product was dissolved
in ether (300 ml) and washed sequentially with 5% HCl,
saturated aqueous KHCO3 (2×), and water. After drying,
the solvent was evaporated leaving 6.1 g (91%) of diac-
etate 14 pure enough for further processing. An analytical
sample was crystallized from methanol, m.p. 103–104 ◦C,
=
1
(C C); 1048 (C–OH). H NMR (200 MHz), δ (ppm): 5.39
(1H, m, W = 12, H-6); 3.55 (1H, m, W = 32, H-3␣); 3.36
(1H, d, J = 5.6, OH); 0.89 (3H, s, 3 × H-18). EI-MS, m/z
(%): 206 (45), 216 (15), 228 (7), 234 (12), 237 (9), 245 (9),
255 (53), 273 (46), 291 (100, M+); 0.5 (M −3), 0.7 (M −2),
4.1 (M − 1), 100.0 (M), 21.0 (M + 1), 2.2 (M + 2); DHEA
for comparison: 203 (46), 213 (20), 228 (18), 231 (12),
237 (9), 242 (6), 255 (59), 270 (37), 288 (100, M+); 0.5
(M −2), 0.7 (M −1), 100.0 (M), 21.2 (M +1), 2.2 (M +2).
Analysis calculated for C19H252H3O2 (291.5): C, 78.30; H,
2
8.65; 2H, 2.07. Found: C, 78.52; H, 9.86 (distorted by H).
2.2.6. Pyridinium 17-oxo[19-2H3]androst-5-en-3β-yl
sulfate (11)
[α] −17 (c 0.64, CHCl3). Ref. [18] gave m.p. 104.5–105 ◦C,
d
[α] − 21 (c 0.6, CHCl3). IR (CHCl3), ν (cm−1): 1727
d
=
=
A solution of d3-DHEA (10) (300 mg, 1.03 mmol) and
sulfur(IV) oxide–pyridine complex (300 mg, 1.88 mmol) in
pyridine (5 ml) was stirred and heated to 70 ◦C for 1 h. The
reaction mixture was then evaporated nearly to dryness, and
hot water (4 ml) was added. The solution was left to at-
tain room temperature and then applied onto a column of
Lichroprep RP-18 (40–63 ␮m, 40 g) prewashed with water.
The column was washed with water, and the product was
eluted with methanol. The fractions containing the sulfate
were collected, and the solvent was evaporated. Crystalliza-
tion from a methanol/ether mixture gave 360 mg (78%) of
(C O, acetate); 1703 (C O, ketone); 1370 (CH3, acetate);
1
1256, 1034 (C–O, acetate). H NMR (200 MHz), δ (ppm):
5.63 (1H, m, W = 12, H-6); 4.64 (1H, m, W = 32, H-3␣);
4.49 (1H, d, J = 11.9, H-19); 3.96 (1H, d, J = 11.9,
H-19ꢀ); 2.12 (3H, s, 3×H-21); 2.03 (6H, s, OAc); 0.65 (3H,
s, 3 × H-18). Analysis calculated for C25H36O5 (416.6): C,
72.08; H, 8.71. Found: C, 72.06; H, 8.86.
2.3.2. 3β-Hydroxy-20-oxopregn-5-en-19-yl acetate (15)
Diacetate 14 (5.6 g, 13.4 mmol) was dissolved in methanol
(500 ml), 2% aqueous KHCO3 (60 ml) was added under
stirring, and the mixture was refluxed and stirred for 2 h
(TLC: benzene/acetone (6:4)). Most of the methanol was
then evaporated on a rotary evaporator, and the mixture
was diluted with water (300 ml) and extracted with ether
(500 ml). The ethereal layer was washed with water (2×),
dried, and the solvent was evaporated, leaving a crude mix-
ture (5.2 g) in which monoacetate 15 prevailed. Column
chromatography on silica gel (70 g) in a mixture of ben-
zene/acetone (from 20:1 to 10:1) yielded 4.4 g (87%) of
sulfate 11, m.p. 188–192 ◦C, [α] + 6 (c 0.4, CHCl3). Ref.
d
[16] gave m.p. 194–195 ◦C for non-deuterated analog. H
1
NMR (400 MHz), δ (ppm): 8.99, 8.49, and 8.01 (5H, pyri-
dinium); 5.42 (1H, m, W = 12, H-6); 4.35 (1H, m, W = 32,
H-3␣); 0.88 (3H, s, 3 × H-18). FAB-MS, m/z: 451 (M + 1).
Analysis calculated for C24H302H3NO5S (450.6): C, 63.97;
H, 6.71; 2H, 1.34; N, 3.11; S, 7.12. Found: C, 63.78; H, 7.63
(distorted by 2H); N, 3.02; S, 7.11.
2.2.7. Sodium 17-oxo[19-2H3]androst-5-en-3β-yl sulfate
(12)
monoacetate 15, m.p. 143–145 ◦C (acetone/hexanes), [α] −
d
20 (c 0.6, CHCl3). IR (CHCl3), ν (cm−1): 3608 (O–H);
=
=
Pyridinium salt 11 (335 mg, 0.74 mmol) was thoroughly
shaken with water (6 ml), and a 10% aqueous NaCl solu-
tion (6 ml) was added in portions. After 1 h of standing in
the refrigerator, the crystals were filtered off, washed with
a small amount of ice cold water and dried. Crude sul-
fate (260 mg) was recrystallized by dissolving in methanol
1726 (C O, acetate); 1701 (C O, ketone); 1367 (CH3, ac-
etate); 1236 (C–O, acetate). For 1H and 13C NMR data, see
Tables 1 and 2. Analysis calculated for C23H34O4 (374.5):
C, 73.76; H, 9.15. Found: C, 73.48; H, 9.25.
From the less polar fractions, diacetate 14 (360 mg, 6%)
was regenerated.

ˇ
I. Cerny´ et al. / Steroids 69 (2004) 161–171
166
2.3.3. 3β,19-Dihydroxypregn-5-en-20-one (16)
2.3.5. 19-Hydroxy-6β-methoxy-3α,5-cyclo-5α,17α-
pregnan-20-one (19)
After further elution with a mixture of benzene/acetone
(1:1), chromatography of a mixture from the preced-
ing experiment gave a sample of diol 16 (120 mg, 3%),
This compound was a side product in the preparation of
cycloderivative 18, from less polar fractions in the final chro-
matography. The yield of 19 was 350 mg (9%, referred to
m.p. 195–196 ◦C (methanol), [α] + 28 (c 0.30, CHCl3).
d
Literature [13] gave m.p. 193–194 ◦C (benzene). IR
15), m.p. 148–149 ◦C. H NMR (400 MHz), δ (ppm): 4.42
1
(CHCl3), ν (cm−1): 3611, 3480 (O–H); 1699 (C O, ke-
(1H, dd, J = 0.9, 11.8, OH); 3.58 (1H, dd, J = 11.4, 11.8,
H-19); 3.38 (3H, s, OCH3); 3.25 (1H, d, J = 11.4, H-19ꢀ);
2.83 (1H, dd, J = 2.7, 8.5, H-17␣); 2.79 (1H, t, J = 2.8,
H-6␣); 2.14 (3H, s, 3 × H-21); 1.02 (3H, s, 3 × H-18); 0.61
(1H, dd, J = 4.0, 5.2, H-4␣); 0.46 (1H, dd, J = 8.2, 5.2,
H-4␤).
=
tone); 1386, 1358 (CH3); 1042 (C–OH). For H and 13C
1
NMR data, see Tables 1 and 2. Analysis calculated for
C21H32O3 (332.5): C, 75.86; H, 9.70. Found: C, 76.01; H,
9.62.
2.3.4. 19-Hydroxy-6β-methoxy-3α,5-cyclo-5α-pregnan-
20-one (18)
2.3.6. 19-Hydroxy-3β-methoxypregn-5-en-20-one (20)
This compound was a side product in the preparation of
cycloderivative 18, from more polar fractions in the final
chromatography. The yield of 20 was 400 mg (10%, referred
Monoacetate 15 (4.3 g, 11.5 mmol) was dissolved in
pyridine (50 ml), the solution was cooled in an ice bath,
and methanesulfonyl chloride (2.7 ml, 35 mmol) was added.
After 1 h (TLC: benzene/acetone (10:1)), the reaction mix-
ture was poured into an ice and water mixture (300 ml),
and crystallization of the product was induced by friction
of a glass stick against the walls of the beaker. Crys-
tals were filtered off, and washed with water. The crude
product was dissolved in benzene (350 ml) and sequen-
tially washed with 5% aqueous citric acid (2×), saturated
aqueous KHCO3 (2×), and water. After drying, the sol-
vent was evaporated, and the residue was dried in vacuo.
The crude mesylate was dissolved in dioxane (25 ml) and
methanol (50 ml), freshly fused potassium acetate (4.0 g,
41 mmol) was added, and the mixture was stirred and
refluxed for 5 h (TLC: benzene/petroleum ether/acetone
(10:10:1)). Methanol was then evaporated on a rotary evap-
orator, water and ice were added (200 ml), and the mixture
was extracted with ether (200 ml). The extract was washed
with water with a small amount of saturated aqueous
NaCl solution (3×), dried, and the solvent was evapo-
rated. A solution of sodium hydroxide (4.8 g, 120 mmol)
in methanol (400 ml) was added to a crude cycloderiva-
tive 17, and the mixture was refluxed and stirred for 1 h
(TLC: benzene/petroleum ether/acetone (10:10:1)). After
cooling, a lump of solid carbon dioxide was added, and
the mixture was concentrated to a small volume. Then
ice-water (200 ml) was added, and the product was ex-
tracted with ether (2× 100 ml). The ethereal layer was
washed with water with small amount of saturated aqueous
NaCl solution (3×), dried, and the solvent was evaporated
to give 3.7 g of a foamy, crude product. Purification by
column chromatography on silica gel (200 ml) in a mix-
ture of benzene/acetone (from 20:1 to 10:1) gave 2.5 g of
cycloderivative 18 (63% from 15). After crystallization
from methanol, 1.7 g of 18 (43% from 15) was obtained,
to 15), m.p. 161–163 ◦C, [α] +24 (c 0.6, CHCl3). 1H NMR
d
(200 MHz), δ (ppm): 5.75 (1H, m, W = 12, H-6); 3.84
(1H, dd, J = 2.1, 11.5, H-19); 3.59 (1H, dd, J = 9.1, 11.5,
H-19ꢀ); 3.36 (3H, s, OCH3); 3.12 (1H, m, W = 32, H-3␣);
2.12 (3H, s, 3 × H-21); 0.67 (3H, s, 3 × H-18).
2.3.7. Methyl 3β-hydroxy-20-oxopregn-5-en-19-oate (23)
Freshly prepared Jones reagent (2.67 M, 9 ml) was
added dropwise to a stirred solution of cycloderivative 18
(1.5 g, 4.33 mmol) in acetone (90 ml) under argon and in
an ice bath. The reaction mixture was stirred and cooled
for 2.5 h (TLC: chloroform/acetone (10:1)). 2-Propanol
(2 ml) was then added, and the mixture was stirred while
cooling for 5 min and then, poured into saturated, aque-
ous KHCO3 (100 ml). Part of the acetone was removed
on a rotary evaporator, additional KHCO3 (100 ml) was
added, and the product was extracted with ethyl acetate
(2× 150 ml). The organic layers were collected, dried,
and the solvent was evaporated. The crude product 21
(1.2 g) was dissolved in acetone (25 ml), 35% perchloric
acid (2.5 ml) was added, and the mixture was stirred at
room temperature for 1 h. After pouring into ice-water
(100 ml), crystalline product was filtered off, washed with
water, and dried under air. Crude 22 (950 mg) was dis-
solved in methanol (30 ml), cooled in an ice bath, and
methylated with ethereal diazomethane solution (to yel-
low color). After 5 min (TLC: chloroform/acetone (10:1)),
several drops of acetic acid were added, and the solu-
tion was evaporated to dryness (1.14 g). Column chro-
matography on silica gel (150 ml) in a mixture of chlo-
roform/acetone (20:1) gave 950 mg of ester 23 (61%).
An analytical sample was crystallized from a mixture
of acetone/petroleum ether, m.p. 175–176 ◦C, [α] − 39
d
m.p. 139–141 ◦C, [α] + 130 (c 0.6, CHCl3). IR (CHCl3),
(c 0.6, CHCl3). IR (CHCl3), ν (cm−1): 3608 (OH);
d
ν (cm−1): 3394 (OH); 3004 (CH2, cyclopropane); 1698
1720 (C O, ester); 1701 (C O, ketone); 1438 (CH3, es-
ter); 1199, 1168 (C–O); 1048 (C–OH). For 1H and 13C
NMR data, see Tables 1 and 2. Analysis calculated for
C22H32O4 (360.5): C, 73.30; H, 8.95. Found: C, 73.14; H,
9.18.
=
=
1
=
(C O); 1435 (CH3, OMe); 1014 (cyclopropane). For H and
13C NMR data, see Tables 1 and 2. Analysis calculated for
C22H34O3 (346.5): C, 76.26; H, 9.89. Found: C, 76.23; H,
10.10.

ˇ
I. Cerny´ et al. / Steroids 69 (2004) 161–171
167
2.3.8. Methyl 20,20-ethylenedioxy-3β-(methoxymethoxy)
pregn-5-en-19-oate (25)
3 × H-21); 0.83 (3H, s, 3 × H-18). Analysis calculated for
C25H382H2O5 (422.6): C, 71.05; H, 9.06; 2H, 0.95. Found:
2
Ethylene glycol (1.4 ml, 25 mmol), triethyl orthoformate
(4.1 ml, 24.6 mmol), and 4-toluenesulfonic acid mono-
hydrate (10 mg) were added to a suspension of ester 23
(900 mg, 2.50 mmol) in benzene (20 ml). The mixture
was stirred at room temperature for 2 h (TLC: petroleum
ether/acetone (3:1)), and during this time the suspension
dissolved. The reaction mixture was poured into ice cooled,
saturated aqueous KHCO3 and extracted with ethyl acetate
(100 ml). The extract was washed sequentially with KHCO3
(2×), water, dried, and the solvent was evaporated. Crude
ketal 24 (1 g) was dissolved in benzene (20 ml), cooled in an
ice bath, and N,N-diisopropylethylamine (5 ml, 28.7 mmol)
and bromomethyl methyl ether (0.6 ml, 7.4 mmol) were
added. The mixture was stirred at room temperature for
2 h (TLC: benzene/acetone (10:1)). Then, the product was
partitioned between ethyl acetate (100 ml) and cold 5%
aqueous citric acid (100 ml). The organic layer was washed
sequentially with citric acid, water, saturated aqueous
KHCO3 (2×), and water, dried, and the solvent was evap-
orated. Crude product (800 mg) in benzene was applied
onto a short column of alumina (30 ml) in benzene and
eluted with a mixture of benzene/acetone (20:1). The yield
of protected ester 25 was 660 mg (59%), m.p. 137–138 ◦C
C, 71.14; H, 9.74 (distorted by H).
2.3.10. 20,20-Ethylenedioxy-3β-(methoxymethoxy)
[19-2H3]pregn-5-ene (28)
Hydroxy derivative 26 (350 mg, 0.83 mmol) was dis-
solved in pyridine (3.5 ml), chilled in an ice bath, and
methanesulfonyl chloride (0.35 ml, 4.52 mmol) was added.
The reaction mixture was left in an ice bath for 1 h. Then, it
was poured onto ice, and the solid product formed was fil-
tered off. Solids were dissolved in ether (30 ml), the solution
was sequentially washed with ice cold 5% aqueous citric
acid (2×), water, saturated aqueous KHCO3 (2×), and wa-
ter. After drying, the solvent was evaporated, and the residue
was coevaporated with benzene (3× 5 ml) and thoroughly
dried in vacuo. Crude mesylate 27 (380 mg) was refluxed
under an argon atmosphere in 1,2-dimethoxyethane (4 ml)
with sodium iodide (400 mg, 2.67 mmol), zinc (400 mg,
6.12 mmol), and deuterium oxide (0.4 ml, 22.11 mmol) for
6 h. The reaction mixture was cooled, diluted with ether,
and filtered through a celite column, which was washed
with ether (ca. 30 ml). The ethereal solution was washed se-
quentially with water, ice cold 5% aqueous citric acid (2×),
water, saturated aqueous KHCO3, 5% aqueous sodium thio-
sulfate pentahydrate, and water. After drying, the solvent
was evaporated. The residue (330 mg) was dissolved in a
minimum of benzene and introduced onto a short column
of aluminum oxide (20 ml). The product was eluted with
a mixture of benzene/ether (100:1). The yield of protected
d3-pregnenolone 28 was 280 mg (83%), m.p. 110–112 ◦C
(methanol), [α] − 73 (c 0.5, CHCl3). IR (CHCl3), ν
d
(cm−1): 1720 (C O, ester); 1675 (C C); 1440 (CH3, es-
ter); 1148, 1103, 1037 (COCOC, methoxymethoxy); 1103,
1055, 950 (ring, dioxolane). 1H NMR (200 MHz), δ (ppm):
5.67 (1H, m, W = 12, H-6); 4.67 (2H, s, O–CH2–O); 3.92
(4H, m, OCH2CH2O); 3.70 (3H, s, COOCH3); 3.44 (1H,
m, W = 30, H-3␣); 3.36 (3H, s, CH3–O–CH2); 1.28 (3H,
s, 3 × H-21); 0.71 (3H, s, 3 × H-18). Analysis calculated
for C26H40O6 (448.6): C, 69.61; H, 8.99. Found: C, 69.40;
H, 9.24.
=
=
(methanol), [α] −32 (c 0.3, CHCl3). IR (CHCl3), ν (cm−1):
d
2
=
2224 (C H3); 1669 (C C); 1147, 1102, 1036 (COCOC,
1
methoxymethoxy); 1102, 1048, 950 (ring, dioxolane). H
NMR (400 MHz), δ (ppm): 5.35 (1H, m, W = 12, H-6);
4.69 (2H, s, O–CH2–O); 3.93 (4H, m, OCH2CH2O); 3.42
(1H, m, W = 32, H-3␣); 3.37 (3H, s, CH3–O–CH2); 1.30
(3H, s, 3 × H-21); 0.78 (3H, s, 3 × H-18). Analysis cal-
2.3.9. 20,20-Ethylenedioxy-3β-(methoxymethoxy)
[19-2H2]pregn-5-en-19-ol (26)
culated for C25H392H3O4 (409.6): C, 73.71; H, 9.60; H,
2
Protected ester 25 (419 mg, 0.93 mmol) in THF (15 ml)
was refluxed with lithium aluminum deuteride (110 mg,
2.62 mmol) under argon for 4 h. The reaction mixture was
then chilled in an ice bath, the remaining deuteride was
destroyed with saturated aqueous Na2SO4, and the mixture
was diluted with ether (15 ml). The solids were filtered
off on a celite column, which was then washed with ether
(50 ml). The resulting solution was sequentially washed
with cold, 5% aqueous citric acid, water, saturated aque-
ous KHCO3 (2×), and water. After drying, the solvents
were evaporated, leaving 380 mg (96%) of hydroxy deriva-
2
1.47. Found: C, 73.75; H, 10.37 (distorted by H).
2.3.11. 3β-Hydroxy[19-2H3]pregn-5-en-20-one (29)
Protected d3-pregnenolone 28 (170 mg, 0.42 mmol) was
stirred in benzene (3 ml) and methanol (3 ml) solution with
35% aqueous HClO4 (0.3 ml) at 50 ◦C for 2 h. The reaction
mixture was cooled, poured into cold saturated aqueous
NaHCO3, and extracted with ethyl acetate (30 ml). The
extract was washed sequentially with saturated aqueous
NaHCO3 and water, dried, and the solvents were evaporated.
The residue was chromatographed on silica gel (20 ml) in
a mixture of benzene/acetone (100:1 to 50:1). The yield of
d3-pregnenolone (29) was 100 mg (75%), m.p. 185–187 ◦C
tive 26, m.p. 132–133 ◦C (methanol), [α] − 26 (c 0.6,
d
CHCl3). IR (CHCl3), ν (cm−1): 3624, 3574 (OH); 2218,
2
=
2115 (C H2); 1666 (C C); 1147, 1103, 1036 (COCOC,
1
methoxymethoxy); 1103, 1051, 950 (ring, dioxolane). H
(methanol), [α] + 25 (c 0.2, ethanol). Ref. [15] gives
d
NMR (200 MHz), δ (ppm): 5.75 (1H, m, W = 12, H-6);
4.68 (2H, s, O–CH2–O); 3.91 (4H, m, OCH2CH2O); 3.47
(1H, m, W = 32, H-3␣); 3.36 (3H, s, CH3O); 1.29 (3H, s,
m.p. 193 ◦C, [α] + 28 (ethanol) for the non-deuterated
d
analog. For H and 13C NMR data, see Tables 1 and 2. IR
1
(CHCl3), ν (cm−1): 3609, 3458 (OH); 2225 (C2H3); 1699

ˇ
I. Cerny´ et al. / Steroids 69 (2004) 161–171
168
=
=
(C O); 1668 (C C); 1368 (CH3); 1358 (CH3CO); 1048
(C–OH). EI-MS, m/z (%): 43 (100), 208 (15), 216 (18),
234 (32), 258 (9), 283 (17), 301 (30), 319 (59, M+); 1.2
(M −3), 0.8 (M −2), 5.4 (M −1), 100.0 (M), 23.1 (M +1),
3.0 (M + 2); PREG for comparison: 43 (100), 205 (15),
213 (14), 231 (23), 255 (7), 283 (18), 298 (20), 316 (37,
M+); 0.2 (M − 2), 0.8 (M − 1), 100.0 (M), 23.2 (M + 1),
2.6 (M + 2). Analysis calculated for C21H292H3O2 (319.5):
C, 78.95; H, 9.15; 2H, 1.89. Found: C, 78.98; H, 10.35
(distorted by 2H).
and 13C NMR data, see Tables 1 and 2. Analysis calculated
for C21H282H3NaO5S (418.5): C, 59.83; H, 6.70; 2H, 1.43;
S, 7.61. Found: C, 59.57; H, 7.49 (distorted by 2H); S, 7.54.
3. Results and discussion
At the initial stage of our synthetic strategy (Scheme 1),
we needed a 19-oate derivative. We reproduced oxidation
experiments of 19-hydroxyandrost-5-ene derivatives with
Jones reagent on 3␤-acetoxy-19-hydroxyandrost-5-en-17-
one, but we obtained only low yields of the corresponding
acid in a mixture with the aldehyde. These results are
comparable to findings in the literature [11]. More promis-
ing were experiments on the 3␣,5-cyclo-5␣-androstane
skeleton (Scheme 2). We used an idea from the patent
literature [10], and we were able to prepare ester 4 from
19-hydroxy-3␣,5-cyclo-5␣-androstan-17-one (1), Ref. [13],
in a three-step procedure (1 → 2 → 3 → 4) in 68% yield.
For the second stage, introduction of two deuterium
atoms through lithium aluminum deuteride reduction of the
ester function in position 19, we needed a suitably pro-
tected derivative. For the ketone in position 17, we chose
an ethylenedioxo group, and the hydroxyl in position 3 was
protected as a methoxymethyl ether. We tried both possible
sequences, i.e. first introducing ether and then ketal and
vice versa, and we found that better yields were obtained by
using ketalization before ether formation (Scheme 3). The
reason for this was that acid catalysis during ketalization
partially cleaved the methoxymethyl group. The protected
derivative 6 was treated with lithium aluminum deuteride in
boiling tetrahydrofuran to give the dideuterated alcohol 7.
The last steps of the synthesis involved preparation of
mesylate 8, its exchange to deuterium, and final depro-
tection of trideutero derivative 9 (Scheme 4). We kept
the conditions for the exchange reaction as anhydrous as
possible. All solids, solvents, and reaction apparatus were
dried before the reaction (see Section 2). After the removal
of both protecting groups in acidic medium, we obtained
d3-DHEA (10), for the deuterium content see discussion
below.
2.3.12. 3β-Hydroxy[19-2H2]pregn-5-en-20-one (30)
A sample from crude mesylate 27 (25 mg) was processed
as described for the preparation of 28 with the use of H2O in-
stead of deuterium oxide and then deprotected as described
for 29. After chromatography, the resulting product (30,
7 mg) was used to obtain NMR spectra for comparison with
the fully deuterated product. For 1H and 13C NMR data, see
Tables 1 and 2. EI-MS, m/z (%): 43 (100), 207 (7), 215 (7),
233 (12), 257 (4), 283 (8), 300 (12), 318 (27, M+).
2.3.13. Pyridinium 20-oxo[19-2H3]pregn-5-en-3β-yl
sulfate (31)
A solution of d3-pregnenolone (29) (250 mg, 0.78 mmol)
and sulfur(IV) oxide–pyridine complex (250 mg, 1.57 mmol)
in pyridine (4 ml) was stirred and heated to 70 ◦C for 1 h. The
reaction mixture was then evaporated nearly to dryness and
hot water (4 ml) was added. The solution was left to attain
room temperature and then, applied onto a column of Lichro-
prep RP-18 (40–63 ␮m, 40 g) prewashed with water. The
column was washed with water, and the product was eluted
with methanol. The fractions containing the sulfate were col-
lected, and the solvent was evaporated. Crystallization from
a methanol-ether mixture gave 260 mg (69%) of sulfate 31,
m.p. 155–160 ◦C, [α] + 20 (c 0.4, CHCl3). Ref. [19] gave
d
m.p. 173–174 ◦C for the non-deuterated analog. IR (CHCl3),
ν (cm−1): +3070, 3020, 1638, 1549, 1490 (pyridinium); 2459,
2
=
2135 (NH ); 2224 (C H3); 1699 (C O); 1271–1175, 1050,
980, 959 (sulfate). For 1H and 13C NMR data, see Tables 1
and 2. Analysis calculated for C26H342H3NO5S (478.7): C,
65.24; H, 7.16; 2H, 1.26; N, 2.93; S, 6.70. Found: C, 65.25;
2
H, 7.97 (distorted by H); N, 2.87; S, 6.69.
2.3.14. Sodium 20-oxo[19-2H3]pregn-5-en-3β-yl sulfate
(32)
Pyridinium salt 31 (160 mg, 0.33 mmol) was thoroughly
shaken with water (3 ml) and a 10% aqueous solution of
NaCl was added in portions. During the addition, the gel-like
solution was changed into a crystalline suspension. After
being kept in refrigerator overnight, the crystals were fil-
tered off, washed with a small amount of ice cold wa-
ter, and dried. The yield of sodium salt 32 was 118 mg
(83%), m.p. 201–204 ◦C, [α] + 19 (c 0.2, ethanol). Ster-
d
aloids gave m.p. 180–182 ◦C, [α] + 21 (ethanol) for the
d
non-deuterated analog. IR (KBr), ν (cm−1): 2220 (C2H3);
Scheme 2. (i) CrO3, H2SO4, acetone; (ii) HClO4, acetone; (iii) CH2N2,
Et2O, MeOH.
1
=
1706 (C O); 1289–1218, 1065, 986, 969 (sulfate). For H

ˇ
I. Cerny´ et al. / Steroids 69 (2004) 161–171
169
Scheme 3. (i) ethylene glycol (EtO)3CH, TsOH, benzene; (ii)
BrCH2OCH3, (I-Pr)2EtN, benzene; (iii) LrAlD4, THF.
Scheme 6. (i) Ac2O, Py; (ii) K2CO3, MeOH; (iii) 1. MsCl, Py, 2. MeOH,
KOAc, dioxane; (iv) NaOH, MeOH.
Scheme 4. (i) MsCl, Py; (ii) NaI, Zn, D2O, DME; (iii) HClO4, benzene,
MeCH.
employing preferential hydrolysis of the 3-O-acetyl group
under mild alkaline conditions [21]. We obtained diol 16
as a side product. Standard procedure then gave 3␣,5-
cycloderivative 17. The next step was the alkaline hydrolysis
of 19-acetate. This step was the most critical in the whole
synthesis because an acetate in this position is relatively
stable, and the pregnane side chain in position 17 partially
isomerizes under the conditions used from 17␤ to 17␣. We
carefully monitored the composition of the reaction mixture
and stopped the reaction when all of starting compound had
reacted. Product 18 was purified by column chromatogra-
phy and crystallization, and the structure was checked by
NMR spectra. In addition, we isolated two side products of
this two-step procedure (15 → 17 → 18), originating from
different steps. 17␣-Isomer 19 came from the second step
and the methoxy derivative 20 from the first step. In spite
of this complication, we obtained the 19-hydroxy derivative
18 in a 43% yield (from 19-acetate 15).
A sulfate was prepared (Scheme 5) by the reaction of 10
with sulfur(VI) oxide–pyridine complex in pyridine [19],
and the resulting pyridinium salt 11 was concentrated on a
reversed phase column and eluted with methanol [20]. Trans-
formation into sodium salt 12 was achieved by treatment
with aqueous sodium chloride [16].
Fewer 19-hydroxy derivatives are known in the preg-
nane series. We started (Scheme 6) from 19-hydroxy-20-
oxopregn-5-en-3␤-yl acetate (13) [7,9], which was acety-
lated to diacetate 14 and transformed into 19-acetate 15,
The rest of the synthesis was done after the sequence cor-
roborated in the androstane series (Scheme 7). The oxidation
of cycloderivative 18 with Jones reagent gave 19-acid 21,
which was transformed without purification into 3-hydroxy-
5-ene derivative 22 and methylated to ester 23. The pro-
tection in position 20 (Scheme 8) gave the ethylenedioxo
derivative 24, and the introduction of the methoxymethyl
group into position 3 gave the fully protected derivative 25.
Scheme 5. (i) P4ASO3, P4; (ii) NaCl, H2O.

ˇ
I. Cerny´ et al. / Steroids 69 (2004) 161–171
170
The sulfate was also prepared by the same method as for
the androstane series. Firstly, we obtained pyridinium salt
29, which was then transformed into sodium salt 30.
The structure of key compounds in the PREG series was
1
confirmed by H and 13C NMR data. A general strategy
based on a combination of 1D and 2D NMR methods was
used for complete structural assignment of all proton and
carbon signals. NMR data are summarized in Tables 1 and 2.
Compounds 13, 15, and 16 were selected as representa-
tives of 19-hydroxypregnenolone derivatives. In comparison
with pregnenolone and pregnenolone acetate, characteristic
signals of H-19 protons in the 1H NMR spectra were appar-
ent in the region of δ = 3.6–3.9 (δ = 4.0–4.5 for acetates).
In the 13C NMR spectra, the carbons mostly influenced were
in positions 1, 5, 6, 10, and 19 (differences in shifts >1 ppm).
Compounds 29–32, where two and/or three protons in
position 19 were replaced with deuterium, showed char-
acteristic changes in the appearance and position of the
Scheme 7. (i) CrO3, H2SO4, acetone; (ii) HClO4, acetone; (iii) CH2N2,
Et2O, MeOH.
Lithium aluminum deuteride reduction yielded the hydroxy
derivative 26, and third deuterium embodiment via mesylate
27 gave 28. The removal of the protecting groups resulted
in 3␤-hydroxy[19-2H3]pregn-5-en-20-one (29), for the deu-
terium content see discussion below. As a sample for com-
parison, we prepared dideuterated PREG 30 using water in
mesylate 27 exchange and the same deprotection reaction as
in the preparation of d3-PREG (29).
1
19-methyl group NMR signal. The H NMR spectrum of
the 19-CHD2 group in 30 gave a characteristic pentet at
δ = 0.976 with 2J(H, D) = 1.8 Hz and the upfield iso-
topic shift (−0.035 ppm in comparison with nondeuterated
PREG). Similarly, its 13C NMR spectrum showed a pentet of
1
the 19-CHD2 group at δ = 18.80 with J(C, D) = 19.3 Hz
and an upfield isotopic shift of −0.57 ppm. The presence
of the 19-CD3 group in compounds 29, 31, and 32 led to
the absence of a corresponding signal in the 1H NMR spec-
tra, while the characteristic heptets at δ = 18.51, 18.42, and
1
17.19 with J(C, D) = 18.7 Hz and upfield isotopic shifts
(−0.86 ppm for 29) were observed in their 13C NMR spec-
tra. The complete absence of proton as well as carbon signals
corresponding to less/deuterated isotopomers in position 19
proved the isotopic purity of the samples at the level higher
than 95%.
In the whole series (13, 15, 16, 18, 23, 29–32), the conser-
vation of the 17␤ configuration of the pregnane side chain
followed from both 1H and 13C NMR spectra (cf. Ref. [22]).
The isotopic abundance estimation was performed for
title compounds d3-DHEA (10) and d3-PREG (29), which
gave suitable EI mass spectra with sufficiently populated
molecular ion. The relative abundances of partially deuter-
ated species were calculated from normalized values of
abundances in the molecular ion cluster (see Section 2).
The values were sequentially corrected to the occurrence
of natural isotopes 13C, 17O, and 18O (correction factors
for M + 1, M + 2; DHEA: 0.211, 0.025; PREG: 0.231,
0.029). After normalization, the abundances d0, d1, d2, d3,
for deuterated DHEA amounted 0.5, 0.6, 3.8, 95.1, and for
PREG 1.1, 0.5, 5.0, 93.4, respectively. The used method,
however, could give only orienting results in our case be-
cause standard DHEA and PREG had some residual M − 2
and M−1 ions (see Section 2). It could be guessed, that non-
deuterated and mono-deuterated species were present up to
1%, dideuterated up to 5%, leaving d3-steroid content about
93–95%. This corresponds to the reported results for the
used labeling methods: the deuteride reduction introduced
Scheme 8. (i) ethylene glycol, (ETO)3CH, TsOH, benzene; (ii)
BrCH2OCH3, (i-Pr)2EtN, benzene; (iii) LiAlD4, THF; (iv) MsCl, Py; (v)
NaI, Zn, D2O, DME; (vi) HClO4, benzene, MeOH; (vii) PyASO3, Py;
(viii) NaCl, H2O.

ˇ
I. Cerny´ et al. / Steroids 69 (2004) 161–171
171
first two deuterium atoms nearly quantitatively, for the sec-
ond method, the label content varies depending on water
present during the reaction [12]. The spectra of preceding
deuterated intermediates in both series were not useful for
estimating of deuterium content due to instability of the
molecular ion caused probably by the presence of dioxolane
ring: all protected derivatives displayed this behavior.
In conclusion, we developed a synthesis of trideutero
derivatives of DHEA and PREG from 19-hydroxy-17-
oxoandrost-5-en-3␤-yl acetate and 19-hydroxy-20-oxopregn-
5-en-3␤-yl acetate, respectively. The approach we used tried
to avoid Jones oxidations of 19-hydroxy group on a 5-ene
skeleton, which gave low yields of mixtures of acids and
aldehydes that were difficult to separate. The oxidation of
the 3␣,5-cycloderivatives proceeded much more smoothly,
and although the whole synthesis was longer, the main
reaction steps produced sufficiently pure intermediates to
be used for preparation of labeled compounds. The synthe-
sis gave deuterated derivatives in 14 steps (yields 8.9 and
7.3%, respectively) with d3-DHEA and d3-PREG content
about 93–95%, sufficient for simple tracing experiments. In
connection with the necessary structural check, data from
fully assigned NMR spectra of selected derivatives in the
PREG series were presented.
[4] Khalsa DS. Integrated medicine and the prevention and reversal of
memory loss. Altern Ther Health Med 1998;4:38–43.
[5] Ladurelle N, Eychenne B, Denton D, Blair-West J, Schumacher M,
Robel P, Baulieu E. Prolonged intracerebroventricular infusion of
neurosteroids affects cognitive performances in the mouse. Brain Res
2000;858:371–9.
[6] Hill M, Paˇr´ızek A, Bicˇ´ıková M, Havl´ıková H, Klak J, Fajt T,
et al. Neuroactive steroids, their precursors, and polar conjugates
during parturition and postpartum in maternal and umbilical blood.
1. Identification and simultaneous determination of pregnanolone
isomers. J Steroid Biochem Mol Biol 2000;75:237–44.
[7] Halpern O, Villoti R, Bowers A. C(19)-substituted hormone
analogues. Chem Ind (London) 1963;39–40.
[8] Kirk DN, Rajagopalan MS, Varley MJ. An improved route to
19-hydroxypregn-4-ene-3,20-dione and synthesis of its [19-2H2]
analogue. J Chem Soc, Perkin Trans 1 1983;2225–7.
[9] Terasawa T, Okada T. Convenient preparative routes to 19-hydroxy,
19-oxo, 19-oic, and 19-nor-deoxycorticosterone. Tetrahedron 1986;
42:537–45.
[10] Sankyo Co. Ltd. (assignee). 6␤-Alkoxy-10␤-carboxy-3␣,5-cyclo
steroids. Neth Appl 6,610,073 (January 20, 1967); Chem Abstr 1968;
68:39949v.
[11] Arunchalam T, Santaniello E, Patel K, Caspi E. Synthesis of
C-19-chiral steroids. Preparation of (19R); (19S) and (19RS)-3␤-
hydroxy[19-2H1,19-3H1]androst-5-en-17-ones. J Chem Soc, Perkin
Trans 1 1987;61–9.
[12] Turecˇek F, Vereš K, Kocˇovský P, Pouzar V, Fajkoš J. Deuterium and
tritium labeling with the zinc-sodium iodide method. J Org Chem
1983;48:2233–7.
[13] Tanabe T, Takasaki R, Sakai K, Hayashi R, Morisawa Y, Hashimoto
T. Steroid series. XVI. The preparation of 3␣,5␣-cyclo-6␤,19-
oxidosteroids and its conversion to 19-oxygenated steroid derivatives.
Chem Pharm Bull 1967;15:15–27.
Acknowledgements
[14] Ciba Ltd. (assignee). 19-Oxygenated steroids. Br 994,747 (July 15,
1960); Chem Abstr 1965;63:13363b.
[15] Budavari, S, O’Neil, MJ, Smith, A, Heckelman PE. 11th Merck
Index, Prasterone and Pregnenolone, Rahway, NJ, 1989. p. 7711 and
7735.
[16] Talbot NB, Ryan J, Wolfe JK. The determination of sodium
dehydroisoandrosterone sulfate in water or urine. J Biol Chem 1943;
148:593–602.
[17] D’Alo F. Salts of dehydroepiandrosterone sulfate. Farmaco (Pavia)
Ed Sci 1956;11:477–83; Chem Abstr 1959;53:11445b.
[18] Ehrenstein M, Dunnenberger M. Investigations on steroids. XXVI.
Synthesis of 19-hydroxy-ꢀ4-androstene-3,17-dione. J Org Chem
1956;21:774–82.
ˇ
We thank Mrs. L. Kloucková for skillful technical assis-
´ˇ
tance. We are indebted to Dr. S. Vašıcková for taking the IR
spectra and Dr. P. Fiedler for interpreting them. Our thanks
are due to Mrs. M. Snopková for the measurements of 200
and 400 MHz 1H NMR spectra. Mass spectra were measured
by the staff of the Laboratory of Mass Spectrometry (Dr. K.
Ubik, Head). Elemental analyses were carried out in the An-
alytical Laboratory (Dr. L. Válková, Head). This work was
carried out as a part of the CAS research project Z4 055 905
and supported by the Grant Agency of the Czech Republic
No. 203/02/1440.
[19] Baxendale D, Kirk DN, Rajagopalan MS, Turner AB. Steroid
conjugates. Part IV. The influence of the cation in borohydride
reduction of ketol sulphates.
6.
J Chem Soc (C) 1971;2563–
References
[20] Kirk DN, Rajagopalan MS. 18-Substituted steroids. Part 12. Synthesis
of aldosterone 21-sulphate, and an improved general procedure
[1] Valleé M, Mayo W, Le Moal M. Role of pregnenolone, dehydro-
epiandrosterone and their sulfate esters on learning and memory in
cognitive aging. Brain Res Rev 2001;37:301–12.
for preparing steroid sulphates.
1987;1339–42.
J Chem Soc, Perkin Trans 1
[2] Flood JF, Morley JE, Roberts E. Memory-enhancing effects in male
mice of pregnenolone and steroids metabolically derived from it.
Proc Natl Acad Sci USA 1992;89:1567–71.
[3] Flood JF, Morley JE, Roberts E. Pregnenolone sulfate enhances post-
training memory processes when injected in very low doses into
limbic system structures: the amygdala is by far the most sensitive.
Proc Natl Acad Sci USA 1995;92:10806–10.
[21] Rajagopalan MS, Smith DSH, Turner AB. Steroid conjugates. Part
III. Syntheses of 19-oxygenated dehydro-epi-androsterone sulphates.
J Chem Soc (C) 1971;646–50.
ˇ
[22] Cerný I, Budeˇš´ınský M, Drašar P, Pouzar V. Construction of
the side-chain in 14␤-androst-5-ene derivatives. Preparation of
14␤-pregnenolone. Collect Czech Chem Commun 1994;59:2691–
704.