Hydrolytic behavior of 5α-hydroxy-11β- and 5β-hydroxy- 11α-substituted 19-norsteroids

Article abstract of DOI:10.1016/j.steroids.2003.07.003

Teutsch G. and Bélanger A. treated 5α,10α epoxides with Grignard-reagents catalyzed by copper(I) ions. The reaction with steroidal epoxides proceeded with complete regio- and stereospecificity, leading exclusively to the 11β-substituted compounds. According to our synthetic strategy, the 5,10 epoxide isomers were not separated; instead, the pure 11β, and in some cases, 11α-substituted molecules were isolated after the conjugate addition of the Grignard-reagents, followed by deketalization and dehydration. Surprisingly, appearance of a third compound was generally observed beside the expected deprotected products, and this compound turned out to have a 3-keto-5(10),9(11) structural unit. Starting from pure 3-ethylenedioxy-5α,10α-epoxy-estr-9(11)-ene-17-one and 3-ethylenedioxy-5β,10β-epoxy-estr-9(11)-ene-17-one, four model compounds were synthesized (11α- and 11β-{4-[1,1-(ethylenedioxy)- ethyl]phenyl}-estra-, as well as 11α- and 11β-cyclohexyl-estra- derivatives) to study the process of deprotection and dehydration. 3-keto-5(10),9(11)-derivatives were found to form after deketalization and dehydration only from 11α-substituted derivatives, while 11β-derivatives resulted in only the expected 3-keto-5,9-diene structure. After observing this remarkable difference between the behavior of 11α-, 11β-substituted isomers we decided to take a closer look at the processes of deketalization and dehydration. In order to carry out the hydrolysis under mild conditions, pyridinium paratoluenesulfonate, a weakly acidic salt, was applied. All the intermediate products observed by TLC were isolated. The outcome of the deprotection and elimination reactions can be rationalized by two factors: conjugation of olefins (with the 3-oxo-group or the 11-phenyl group) and orientation of groups to be eliminated.

Full text of DOI:10.1016/j.steroids.2003.07.003

Steroids 68 (2003) 739–749
Hydrolytic behavior of 5␣-hydroxy-11␤- and 5␤-hydroxy-
11␣-substituted 19-norsteroids
Erzsébet Francsics-Czinege∗, Zoltán Tuba, Csaba Molnár, Judit Horváth, János Csörgei,
György Visky, Gábor Balogh, Marianna Mák, Béla Hegedu˝s, Mária Magyari, János Horváth
Department of Steroid Research, Gedeon Richter Ltd., P.O. Box 27, Budapest, 1475, Hungary
Received 28 April 2003; received in revised form 8 July 2003; accepted 21 July 2003
Abstract
TeutschG. andBélangerA. treated5␣,10␣epoxideswithGrignard-reagentscatalyzedbycopper(I)ions. Thereactionwithsteroidalepox-
ides proceeded with complete regio- and stereospecificity, leading exclusively to the 11␤-substituted compounds. According to our synthetic
strategy, the 5,10 epoxide isomers were not separated; instead, the pure 11␤, and in some cases, 11␣-substituted molecules were isolated after
the conjugate addition of the Grignard-reagents, followed by deketalization and dehydration. Surprisingly, appearance of a third compound
was generally observed beside the expected deprotected products, and this compound turned out to have a 3-keto-5(10),9(11) structural unit.
Starting from pure 3-ethylenedioxy-5␣,10␣-epoxy-estr-9(11)-ene-17-one and 3-ethylenedioxy-5␤,10␤-epoxy-estr-9(11)-ene-17-one, four
modelcompoundsweresynthesized(11␣-and11␤-{4-[1,1-(ethylenedioxy)-ethyl]phenyl}-estra-, aswellas11␣-and11␤-cyclohexyl-estra-
derivatives) to study the process of deprotection and dehydration. 3-keto-5(10),9(11)-derivatives were found to form after deketalization
and dehydration only from 11␣-substituted derivatives, while 11␤-derivatives resulted in only the expected 3-keto-5,9-diene structure.
After observing this remarkable difference between the behavior of 11␣-, 11␤-substituted isomers we decided to take a closer look at the
processes of deketalization and dehydration. In order to carry out the hydrolysis under mild conditions, pyridinium paratoluenesulfonate, a
weakly acidic salt, was applied. All the intermediate products observed by TLC were isolated. The outcome of the deprotection and elim-
ination reactions can be rationalized by two factors: conjugation of olefins (with the 3-oxo-group or the 11-phenyl group) and orientation
of groups to be eliminated.
© 2003 Elsevier Inc. All rights reserved.
Keywords: Steroid; Hydrolysis; 11␣-Substituted steroids; 11␤-Substituted steroids; 11-Substituted-5(10),9(11) steroids
1. Introduction
of the conjugate addition of Grignard-reagent into epox-
ides, Teutsch proved that 5␤,10␤ epoxides afforded 11␣-
substituted compounds [6].
The initial breakthrough in finding compounds with
progesterone antagonist activity was purely chemical,
with the discovery of a general access to 11␤-substituted
19-norsteroids. Teutsch G. and Bélanger A. treated 5␣,10␣
epoxides with Grignard-reagent catalyzed by copper(I) ions
[1]. The reaction with steroidal epoxides proceeded with
complete regio- and stereospecificity, leading exclusively
to the 11␤-substituted compounds (2). It is well-known
that compounds of type 2 can be deketalized and con-
comitantly dehydrated by acids yielding molecules with a
3-keto-4,9-diene-system (3) (Scheme 1).
According to our synthetic strategy, the new 5,10 epoxide
isomers were not separated; instead, the pure 11␤, and in
some cases, 11␣-substituted molecules were isolated after
the conjugate addition of the Grignard-reagent, deketaliza-
tion, and dehydration. Surprisingly, the appearance of a third
compound was generally observed besides the expected de-
protected products. In this article, we describe two examples
of the isolated compounds and give an explanation for our
results by examining the dehydration and deketalization of
model compounds.
A great number of 11␤-substituted compounds with anti-
progestational and/or antiglucocorticoid activity were syn-
thesized by Teutsch’s method [2–5]. To complete the study
2. Experimental
∗
Melting points were determined on a BÜCHI-510 appara-
tus and are uncorrected. The NMR spectra were recorded on
Corresponding author. Tel.: +36-1431-5414; fax: +36-1260-0349.
E-mail address: e.francsics@richter.hu (E. Francsics-Czinege).
0039-128X/$ – see front matter © 2003 Elsevier Inc. All rights reserved.
doi:10.1016/j.steroids.2003.07.003

740
E. Francsics-Czinege et al. / Steroids 68 (2003) 739–749
OBz
OBz
OBz
R
R
H
H
H
H+
R-Mg-Br, Cu(I)Cl
O
H
H
H
O
O
O
1
O
O
OH
2
3
R= ethyl, phenyl, etc.
Scheme 1.
either a Varian UnityINOVA-300 or a Varian UnityINOVA-500
spectrometer, applying CDCl3 or DMSO-d6 as a solvent. In-
frared spectra were recorded on a Perkin–Elmer Spectrum
solvent was removed in vacuo, and the residue was crys-
tallized from diisopropyl ether to give a mixture of 5␣,10␣
and 5␤,10␤ isomers of the epoxides. The epoxide ratio was
determined by HPLC and by 1H NMR. Yield: 10.4 g (89%);
5␣,10␣/5␤,10␤ epoxide ratio 57.9/42.1 (HPLC); MS: (m/z)
344 (M+); IR: 2967, 2938, 2878, 1730, 1637, 1117, 1072,
1057, 1006 cm−1. Analysis calculated for C21H28O4: C
73.22, H 8.19; found: C 73.09, H 8.12. 1H NMR {300 MHz,
CDCl3 (TMS), δ (ppm), 59:41 rates mixture of two iso-
mers (5␣,10␣-epoxy/5␤,10␤-epoxy)}: 0.80/0.76 (3H, t,
–CH2–CH3); 3.83–4.00 (4H, m, ethylenedioxy); 6.03/5.83
(1H, m, H-11).
1000 FT-IR instrument with spectral resolution of 4 cm−1
,
using a KBr matrix. Electron ionization mass spectral anal-
yses were carried out by a VG-TRIO-2 quadrupole mass
spectrometer at 70 eV with direct introduction and at an ion
source temperature of 250 ◦C. The “fast atom bombardment”
(FAB) and the “daughter ion” spectra were performed with
a Finnigan MAT 95SQ hybrid tandem mass spectrometer.
The Cs+ gun was used at 20 kV, and the matrix applied was
glycerin or 3-nitrobenzyl alcohol. High resolution mass mea-
surements (HRMS) were performed by Finnigan MAT 95SQ
mass spectrometer, and the accuracy of the peak match-
ing was 1 ppm. Column chromatography was performed
on 63–230 ␮M silica gel obtained from Merck Co, Darm-
stadt, Germany. TLC analyses were carried out on silica
gel 60 F254 aluminum sheets (20 cm × 20 cm with 250 ␮M
layer). Analytical HPLC was conducted using a reverse
phase column Nucleosil 120 C-18 7 ␮m (250 × 4) and ace-
tonitrile:methanol:water 35:35:30 as a mobile phase, the UV
detector (SPD-6A) was set at 205, 260 or 320 nm.
All reactions were carried out under a nitrogen atmo-
sphere. Most chemicals and solvents were analytical grade
and used without further purification.
2-Prop-2-ynyloxy-tetrahydro-pyran was prepared ac-
cording to the procedure of Jones et al. [7] 4-bromo-
acetophenone ethylene ketal was synthesized according to
Neville’s method [8]. Pyridinium p-toluenesulfonate was
synthesized starting from pyridine and p-toluenesulfonic
acid [9].
2.2. 3-Ethylenedioxy-5,10-epoxy-17β-hydroxy-17α-
[3-(tetrahydro-2H-pyran-2-yloxy)-1-propynyl]-18a-
homo-estr-9(11)-ene (7)
A 1.65 mol solution of butyl lithium in n-hexane
(550 ml, 0.88 mol butyl lithium) was added to a solu-
tion of 2-[(prop-2-ynyl)oxy]-tetrahydropyran (6) (141 ml,
0.97 mol) in dry tetrahydrofuran (650 ml), and the mixture
was stirred for 30 min at the same temperature. Compound
5 (0.15 mol), dissolved in dry tetrahydrofuran (80 ml), was
slowly added at 0 ◦C, and the solution was stirred for
30 min, and then poured into saturated aqueous ammonium
chloride solution (730 ml). After separation, the aqueous
fraction was extracted with tetrahydrofuran (2 × 250 ml).
The combined organic fractions were dried over sodium
sulfate and concentrated in vacuo to give an oil. The excess
reagent was removed via column chromatography (in cyclo-
hexane), and after evaporating of the solvent in vacuo, 7 was
retrieved as a foam. Yield: 59.5 g (85%); 5␣,10␣/5␤,10␤
epoxide ratio 63.9/36.1 (HPLC); IR: 3450, 2940, 2878,
2237, 1643, 1119, 1065, 1024, 970, 902 cm−1. HPLC as-
say: 97.9%; HRMS: calculated for C29H39O5 [MH-18]+:
467.2798, found: 467.2808. 1H NMR {500 MHz, CDCl3
(TMS), δ (ppm), 62:38 ratio for rates mixture of 2 iso-
mers (5␣,10␣-epoxy/5␤,10␤-epoxy)}: 0.96/0.90 (3H, t,
–CH2–CH3); 3.47–4.38 (8H, m, –O–CH2–); 4.77 (1H, m,
–O–CH–O–); 6.08/5.90 (1H, m, H-11).
2.1. 3-Ethylenedioxy-5,10-epoxy-18a-homo-estr-9(11)-
ene-17-one (5)
Hexachloroacetone (1.2 ml, 8 mmol) and 70% aqueous
hydrogen peroxide (6.9 ml, 0.18 mol) was added to a solu-
tion of 4 (11.2 g, 0.034 mol) in methylene chloride (75 ml)
at 0 ◦C. The reaction mixture was stirred for 3 h at the same
temperature. Then, it was poured into a mixture of methy-
lene chloride (340 ml) and ice (160 g). A solution of sodium
thiosulphate (31.6 g, 0.2 mol) in water (140 ml) was added
dropwise to the mixture to destroy the excess of hydrogen
peroxide. After separation, the organic fraction was washed
with water (2 × 100 ml) and dried on sodium sulphate. The
2.2.1. Procedure for synthesis of compounds
10a, 11, and 12a
Under anhydrous conditions, dry tetrahydrofuran (6 ml)
was added to magnesium (3.14 g, 0.13 mol). A portion

E. Francsics-Czinege et al. / Steroids 68 (2003) 739–749
741
(8 ml) of a solution of 4-bromo-N,N-dimethylaniline (23.8 g,
0.119 mol) in dry tetrahydrofuran (94 ml) and a drop of di-
bromoethane was added at 50 ◦C. After evidence of reaction
was observed, the entire amount of the reagent was added
dropwise at 15–20 ◦C. The Grignard-reagent was added
dropwise to a suspension of 7 (11.6 g, 0.024 mol) and cop-
per (I) chloride (0.5 g, 0.005 mol) in dry methylene chloride
(60 ml) at 10–15 ◦C. The reaction mixture was stirred for
2 h and then, it was poured into 10% ammonium chloride
solution (360 ml) and extracted with methylene chloride
(3 × 50 ml). The combined organic fractions were washed
with water (4 × 80 ml), dried over sodium sulfate, filtered,
and concentrated in vacuo to give a dark oil (40 g). Purifi-
cation via column chromatography (cyclohexane) gave the
product in the form of a foam, which was hydrolyzed.
1607, 1517, 1071, 1017, 843, 824. Analysis calculated for
C30H37NO3: C 78.39, H 8.11, N 3.06; found: C 78.52, H
8.07, N 3.01. 1H NMR {500 MHz, CDCl3 (TMS), δ (ppm)}:
1.11 (3H, t, –CH2–CH3), 2.95 (6H, s, –N–CH3), 4.23 (2H,
s, CH2–OH), 6.68 (2H, br, H-3ꢀ,5ꢀ), 7.09 (2H, br, H-2ꢀ,6ꢀ).
2.2.2. Procedure for synthesis of compounds 10b and 12b
Under anhydrous conditions, dry tetrahydrofuran (3 ml)
was added to magnesium (1.25 g, 0.052 mol). A portion
(4 ml) of a solution of 4-bromoacetophenone ethylene ke-
tal (12.2 g, 0.05 mol) in dry tetrahydrofuran (50 ml) and a
drop of dibromoethane was added at 50 ◦C. After evidence
of reaction was observed, the entire amount of the reagent
was added dropwise. The reaction mixture was stirred for
an additional 2 h while it was cooling to room temperature.
The mixture was then added dropwise to a suspension of 7
(4.67 g, 0.01 mol) and copper (I) chloride (0.3 g, 0.003 mol)
in dry methylene chloride (30 ml) at 10–15 ◦C. The reac-
tion mixture was stirred for 1 h, and then it was poured into
10% ammonium chloride (70 ml) solution and extracted with
methylene chloride (3×40 ml). The combined organic frac-
tions were washed with water (4×70 ml), dried over sodium
sulfate, filtered, and concentrated in vacuo to give a yellow
oil (9.7 g). Purification via column chromatography (cyclo-
hexane/acetone 9:1) gave a yellowish oil (4.3 g), which was
hydrolyzed.
2.2.1.1. Hydrolysis. The protected product (0.01 mol) was
added to a solution of potassium hydrogensulfate (5.8 g,
0.042 mol) in water (46 ml) under a nitrogen atmosphere.
The reaction mixture was stirred for 2 h. After neutralizing
with 10% potassium hydroxide solution, the mixture was
diluted with water (20 ml) and extracted with methylene
chloride (3 × 40 ml). The organic fractions were washed
with water (2 × 16 ml), dried over sodium sulfate, and
concentrated in vacuo. The oily residue was crystallized in
acetonitrile to give 10a. The acetonitrile mother liquor was
concentrated in vacuo, and 11 as well as 12a were isolated
by column chromatography (cyclohexane–acetone 9:1).
2.2.2.1. Hydrolysis. 2N hydrochloric acid (12 ml) was
added to a solution of the mixture of protected product
(4.14 g, 0.009 mol) in methanol (90 ml). The reaction mix-
ture was stirred for 20 min. After neutralizing with 5%
sodium hydrogencarbonate solution, the mixture was di-
luted with water (50 ml) and extracted with methylene
chloride (3 × 50 ml). The organic fractions were washed
with water (2 × 40 ml), dried over sodium sulfate, and
concentrated in vacuo. The oily residue was crystallized in
acetonitrile to give 10b, while 12b was isolated by column
chromatography (methylene chloride/ethyl acetate 1:1).
11β-(4ꢀ-dimethyl-amino-phenyl)-17β-hydroxy-17-(3-hy-
droxy-1-propynyl)-18a-homo-estra-4,9-diene-3-one (10a).
Yield: 1.25 g (25%); mp: 127–131 ◦C; MS: (m/z) 459 (M+);
IR: 3363, 2940, 2858, 1691, 1593, 1515, 1445, 1308,
1214, 1135, 1044, 822, 611 cm−1. Analysis calculated for
C30H37NO3: C 78.39, H 8.11, N 3.06; found: C 78.51, H
8.03, N 3.11. 1H NMR {500 MHz, DMSO-d6 (TMS), δ
(ppm)}: 0.22 (3H, t, –CH2–CH3), 2.83 (6H, s, –N–CH3),
4.13 (2H, d, –CH2–OH), 4.31 (1H, m, H-11), 5.10 (1H, t,
–CH2–OH), 5.16 (1H, s, 17-OH), 5.61 (1H, br, H-4), 6.62
(2H, m, H-3ꢀ,5ꢀ), 7.03 (2H, m, H-2ꢀ,6ꢀ).
11β-(4ꢀ-Acetyl-phenyl)-17β-hydroxy-17-(3-hydroxy-1-
propynyl)-18a-homo-estra-4,9-diene-3-one (10b). Yield:
1.3 g (24%); mp: 132–34 ◦C; IR: 2914, 1701, 1679, 1579,
1α-(4ꢀ-Dimethyl-amino-phenyl)-17β-hydroxy-17-(3-
hydroxy-1-propynyl)-18a-homo-estra-4,9-diene-3-one (11).
Yield: 0.58 g (12%); mp: 167–72 ◦C; MS: (m/z) 459 (M+);
IR: 3369, 3298, 2931, 2873, 1637, 1617, 1597, 1068,
1019, 812. Analysis calculated for C30H37NO3: C 78.39,
H 8.11, N 3.06; found: C 78.15, H 8.14, N 3.14. 1H
NMR {300 MHz, CDCl3 (TMS), δ (ppm)}: 1.11 (3H, t,
–CH2–CH3), 2.91 (6H, s, –N–CH3), 3.72 (1H, m, H-11),
4.61 (2H, s, CH2–OH), 5.62 (1H, br, H-4), 6.65 (2H, m,
H-3ꢀ,5ꢀ), 6.96 (2H, m, H-2ꢀ,6ꢀ).
1597, 1400, 1358, 1265, 1179, 1073, 951, 837, 598 cm−1
;
MS: (m/z) 458 (M+). Analysis calculated for C30H34O4: C
78.57, H 7.47; found: C 78.38, H 7.41. 1H NMR {300 MHz,
CDCl3 (TMS), δ (ppm)}: 0.24 (3H, t, –CH2–CH3), 2.58
(3H, s, –CO–CH3), 4.36 (2H, s, CH2–OH), 4.45 (1H, m,
H-11), 5.76 (1H, br, H-4), 7.30 (2H, m, H-2ꢀ,6ꢀ), 7.87 (2H,
m, H-3ꢀ,5ꢀ).
11-(4ꢀ-Acetyl-phenyl)-17β-hydroxy-17-(3-hydroxy-1-
propynyl)-18a-homo-estra-5(10),9(11)-diene-3-one (12b).
Yield: 0.8 g (17%); mp: 197–201 ◦C; MS: (m/z) 458 (M+);
IR: 3480, 3320, 2250, 1672, 1633, 1598, 1271, 1025, 837.
Analysis calculated for C30H34O4: C 78.57, H 7.47; found:
C 78.76, H 7.39. 1H NMR {500 MHz, CDCl3 (TMS), δ
11-(4ꢀ-Dimethyl-amino-phenyl)-17β-hydroxy-17-(3-hy-
droxy-1-propynyl)-18a-homo-estra-5(10),9(11)-diene-3-one
(12a). Yield: 0.87 g (18%); mp: 203–206 ◦C; MS: (m/z)
459 (M+); IR: 3411, 3361, 3314, 2930, 2877, 1701,

742
E. Francsics-Czinege et al. / Steroids 68 (2003) 739–749
(ppm)}: 1.14 (3H, t, –CH2–CH3), 2.59 (3H, s, –CO–CH3),
2.79 & 2.86 (2H, d & d, H-4), 4.23 (2H, s, CH2-OH), 7.32
(2H, br, H-2ꢀ,6ꢀ), 7.88 (2H, m, H-3ꢀ,5ꢀ).
action mixture was stirred at room temperature for 20 min.
Meanwhile, the starting material dissolved, and then, the
solution turned into a suspension. The crystals were fil-
tered off, washed with methanol and water, and dried in
vacuo over potassium hydroxide to yield 17a (0.59 g, 76%);
mp: 180–82 ◦C; MS: (m/z) 388 (M+); IR: 2948, 2873,
1731, 1675, 1657, 1600, 1563, 839. Analysis calculated for
2.3. 3-Ethylenedioxy-5α-hydroxy-11β-{4ꢀ-[1ꢀ,1ꢀ-
(ethylenedioxy)-ethyl]phenyl}-estra-9-ene-17-one (16a)
1
Under anhydrous conditions, dry tetrahydrofuran (3 ml)
was added to magnesium (1.25 g, 0.052 mol). A portion
(4 ml) of a solution of 4-bromoacetophenone ethylene ke-
tal (12.2 g, 0.05 mol) in dry tetrahydrofuran (50 ml) and a
drop of dibromoethane was added at 50 ◦C. After evidence
of reaction was observed, the entire amount of the reagent
was added dropwise. The reaction mixture was stirred for
an additional 2 h while it was cooling to room tempera-
ture. The mixture was then added dropwise to a suspen-
sion of 14 (3.14 g, 0.01 mol) and copper (I) chloride (0.3 g,
0.003 mol) in dry methylene chloride (30 ml) at 0 ◦C. The
reaction mixture was stirred for 1 h, and then, it was poured
into 10% ammonium chloride (70 ml) solution and extracted
with methylene chloride (3 × 40 ml). The combined organic
fractions were washed with water (4 × 70 ml), dried over
sodium sulfate, filtered, and concentrated in vacuo to give a
yellow oil (9.7 g). Purification via column chromatography
(methylene chloride–ethyl acetate 1:1) followed by tritura-
tion in diisopropyl ether gave 16a. Yield: 2.56 g (52%); mp:
183–88 ◦C (lit. 190–91 ◦C); MS: (m/z) 494 (M+); IR: 3561,
3533, 2935, 2890, 1733, 1691, 1610, 1169, 1125, 1084,
1045, 950, 831, 819. Analysis calculated for C30H38O6: C
72.85, H 7.74; found: C 72.68, H 7.60. 1H NMR {500 MHz,
CDCl3 (TMS), δ (ppm)}: 0.47 (3H, s, H-18), 1.63 (3H,
s, –CH3), 3.77 & 4.02 (4H, m, ethylenedioxy), 3.89–4.06
(4H, m, 3-ethylenedioxy), 4.31 (1H, m, H-11), 4.34 (1H, br,
5-OH), 7.19 (2H, m, H-2ꢀ,6ꢀ), 7.35 (2H, m, H-3ꢀ,5ꢀ).
C26H28O3: C 80.38, H 7.26; found: C 80.21, H 7.19. H
NMR {500 MHz, CDCl3 (TMS), δ (ppm)}: 0.54 (3H, s,
H-18), 2.58 (3H, s, –CO–CH3), 4.48 (1H, m, H-11), 5.82
(1H, br, H-4), 7.32 (2H, m, H-2ꢀ,6ꢀ), 7.89 (2H, m, H-3ꢀ,5ꢀ).
2.6. 11β-Cyclohexyl-estra-4,9-diene-3,17-dione (17b)
17b was synthesized from 16b (0.8 g, 0.002 mol) ac-
cording to the procedure given for synthesis of 17a. Yield:
0.74 g (85%); mp: 205–207 ◦C; MS: (m/z) 352 (M+); IR:
2931, 2848, 1735, 1658, 1605, 1441, 1383, 1242, 1033, 862.
Analysis calculated for C24H32O2: C 81.77, H 9.15; found:
C 81.91, H 9.08. 1H NMR {500 MHz, CDCl3 (TMS), δ
(ppm)}: 1.10 (3H, s, H-18), 1.52 (1H, m, H-1ꢀ), 2.72 (1H,
m, H-11), 5.70 (1H, br, H-4).
2.7. 3-Ethylenedioxy-5β-hydroxy-11α-{4ꢀ-[1ꢀ,1ꢀ-
(ethylenedioxy)-ethyl]phenyl}-estra-9-ene-17-one (19a)
19a was synthesized from 18 (3.14 g, 0.01 mol) according
to the procedure given for synthesis of 16a. The reaction
time of 18 with the Grignard-reagent 8b was 4 h.
Yield: 1.45 g (30%); mp: 160–62 ◦C; MS: (m/z) 478
([M-18]+); IR: 3507, 2932, 2876, 1731, 1681, 1660, 1607,
1188, 1160, 1115, 1039, 945, 844, 826. Analysis calculated
for C30H38O6: C 72.85, H 7.74; found: C 73.02, H 7.81.
1H NMR {500 MHz, CDCl3 (TMS), δ (ppm)}: 0.98 (3H,
s, H-18), 1.62 (3H, s, –CH3), 3.69–4.05 (8H, m, ethylene-
dioxy), 3.88 (1H, m, H-11), 4.53 (1H, br, 5-OH), 7.07 (2H,
m, H-2ꢀ,6ꢀ), 7.35 (2H, m, H-3ꢀ,5ꢀ).
2.4. 3-Ethylenedioxy-5α-hydroxy-11β-cyclohexyl-estra-
9-ene-17-one (16b)
16b was synthesized from 14 (3.14 g, 0.01 mol) accord-
ing to the procedure given for the synthesis of 16a. The
Grignard-reagent (14) was prepared from cyclohexyl bro-
mide (6.2 ml, 0.05 mol) without dibromoethane. The prod-
uct was isolated by crystallization from diisopropyl ether to
give 1.70 g (40%) of 16b; mp: 152–154 ◦C; MS: (m/z) 414
(M+); IR: 3519, 2930, 1739, 1447, 1182, 1138, 1107, 978,
840. Analysis calculated for C26H38O4: C 75.32, H 9.24;
found: C 75.48, H 9.31. 1H NMR {500 MHz, CDCl3 (TMS),
δ (ppm)}: 1.03 (3H, s, H-18), 1.38 (1H, m, H-1ꢀ), 2.59 (1H,
m, H-11), 3.94–4.08 (4H, m, ethylenedioxy), 4.34 (1H, d,
5-OH).
2.8. 3-Ethylenedioxy-5β-hydroxy-11α-cyclohexyl-
estra-9-ene-17-one (19b)
19b was synthesized from 18 (3.14 g, 0.01 mol) according
to the procedure given for the synthesis of 16b. The reaction
time of 18 with the Grignard-reagent 15 was 4 h. Purification
was carried out by column chromatography (n-hexane–ethyl
acetate 1:1) to yield 19b (2.14 g, 52%); mp: 184–85 ◦C; MS:
(m/z) 396 ([M-18]+); IR: 3500, 2924, 1736, 1449, 1269,
1187, 1105, 993, 839. Analysis calculated for C26H38O4: C
75.32, H 9.24; found: C 75.53, H 9.18. 1H NMR {500 MHz,
CDCl3 (TMS), δ (ppm)}: 0.86 (3H, s, H-18), 1.51 (1H,
m, H-1ꢀ), 2.78 (1H, m, H-11), 3.93–4.07 (4H, m, ethylene-
dioxy), 4.47 (1H, s, 5-OH).
2.5. 11β-(4ꢀ-Acetyl-phenyl)-estra-4,9-diene-
3,17-dione (17a)
2.8.1. Procedure for synthesis of 20a and 21a
2 N hydrochloric acid (1 ml) was added to a suspension
of 16a (1.0 g, 0.002 mol) in methanol (10 ml) and the re-
2N hydrochloric acid (1 ml) was added to a suspension of
19a (1.0 g, 0.002 mol) in methanol (10 ml) and the reaction

E. Francsics-Czinege et al. / Steroids 68 (2003) 739–749
743
mixture was stirred at room temperature for 30 min. Af-
ter neutralizing with 5% aqueous sodium hydrogencarbon-
ate solution, the mixture was diluted with water (50 ml)
and extracted with methylene chloride (3 × 50 ml). The or-
ganic fractions were washed with water (3 × 35 ml), dried
over sodium sulfate, and concentrated in vacuo. The re-
sulted oily residue was purified by column chromatography
(n-hexane–ethyl acetate 1:1) to yield 20a and 21a as white
crystals.
isolated when they reached their maximum concentration
in the reaction mixture.
2.8.3.1. Work-up and isolation. The reaction mixture was
poured into 5% sodium hydrogencarbonate (100 ml), and
after stirring for 30 min, the precipitate was collected on a
filter washed with water and dried over potassium hydroxide.
The crude product was purified by column chromatography
(n-hexane–ethyl acetate 1:1).
2.9. 3-Ethylenedioxy-5α-ethoxy-11β-{4ꢀ-[1ꢀ,1ꢀ-
(ethylenedioxy)-ethyl]phenyl}-estr-9-ene-17-one (21)
2.8.1.1. 11α-(4ꢀ-Acetyl-phenyl)-estra-4,9-diene-3,17-dione
(20a). Yield: 0.20 g (26%); mp: 176–178 ◦C; MS: (m/z)
388 (M+); IR: 2932, 2874, 1738, 1680, 1660, 1603, 1582,
831. Analysis calculated for C26H28O3: C 80.38, H 7.26;
found: C 80.57, H 7.19. 1H NMR {500 MHz, CDCl3 (TMS),
δ (ppm)}: 1.04 (3H, s, H-18), 2.57 (3H, s, –CO–CH3), 4.18
(1H, m, H-11), 5.73 (1H, br, H-4), 7.17 (2H, m, H-2ꢀ,6ꢀ),
7.87 (2H, m, H-3ꢀ,5ꢀ).
21 was isolated after working up the reaction mixture af-
ter 20 min. Yield: 0.26 g (27%); IR: 1739, 1682, 1610, 1245,
1197, 1123, 1091, 1040, 833. HPLC: 97.8%; HRMS: calcu-
lated for [M-46]+ C30H36O5 476.25627, found: 476.25608.
1H NMR {500 MHz, CDCl3 (TMS), δ (ppm)}: 0.46 (3H, s,
H-18), 1.21 (3H, t, –CH2–CH3), 1.63 (3H, s, –CH3), 3.35
& 3.53 (2H, m, –CH2–CH3), 3.73–4.05 (8H, m, ethylene-
dioxy), 4.32 (1H, m, H-11), 7.19 (2H, m, H-2ꢀ,6ꢀ), 7.35 (2H,
m, H-3ꢀ,5ꢀ).
2.8.1.2. 11-(4ꢀ-Acetyl-phenyl)-estra-5(10),9(11)-diene-3,17-
dione (21a). Yield: 0.42 g (54%); mp: 114–116 ◦C; MS:
(m/z) 388 (M+); IR: 2919, 2856, 1738, 1717, 1681, 1601,
1557, 839. Analysis calculated for C26H28O3: C 80.38,
H 7.26; found: C 80.55, H 7.21. 1H NMR {500 MHz,
CDCl3 (TMS), δ (ppm)}: 1.05 (3H, s, H-18), 2.60 (3H, s,
–CO–CH3), 2.80 & 2.89 (2H, d & d, H-4), 7.29 (2H, m,
H-2ꢀ,6ꢀ), 7.88 (2H, m, H-3ꢀ,5ꢀ).
2.10. 3-Ethylenedioxy-5α-ethoxy-11β-(4ꢀ-acetyl-phenyl)-
estr-9-ene-17-one (23)
23 was isolated by working up the reaction mixture after
60 min. Yield: 0.17 g (20%); IR: 1739, 1691, 1603, 1270,
1186, 1120, 1063, 832. HPLC: 96.9%; HRMS: calculated for
[M-46]+ C28H32O4 432.23006, found: 432.22997. 1H NMR
{500 MHz, CDCl3 (TMS), δ (ppm)}: 0.47 (3H, s, H-18),
1.21 (3H, t, –CH2–CH3), 2.57 (3H, s, –CO–CH3), 3.36
& 3.54 (2H, m, –CH2–CH3), 3.74–3.97 (4H, m, ethylene-
dioxy), 4.38 (1H, m, H-11), 7.35 (2H, m, H-2ꢀ,6ꢀ), 7.87 (2H,
m, H-3ꢀ,5ꢀ).
2.8.2. Procedure for synthesis of 20b and 21b
20b and 21b were prepared from 19b (0.8 g, 0.002 mol)
according to the procedure given for the synthesis of 20a
and 21a. The reaction mixture was stirred for 40 min.
2.8.2.1. 11α-Cyclohexyl-estra-4,9-diene-3,17-dione (20b).
20b was isolated as a white foam (0.36 g, 52%); IR: 2926,
2950, 1739, 1660, 1590, 1449, 1373, 1219, 1064, 873.
HPLC: 97.4%; HRMS: calculated for C24H32O2 352.24023,
2.11. 3-Ethylenedioxy-5α-hydroxy-11β-(4ꢀ-acetyl-phenyl)-
estr-9-ene-17-one (24)
1
found: 352.24045. H NMR {500 MHz, CDCl3 (TMS), δ
(ppm)}: 0.90 (3H, s, H-18), 1.62 (1H, m, H-1ꢀ), 3.06 (1H,
24 was prepared from 16a in 90% ethanol according to
the procedure given for 22 by working up the reaction mix-
ture after 3.5 h. TLC monitoring and column chromatogra-
phy was carried out in n-hexane–methylene chloride 1:1.
Yield: 0.39 g (40%); IR: 3535, 1678, 1601, 1565, 1269,
1134, 1121, 1051, 838. HPLC: 97.8%; HRMS: calculated
m, H-11), 5.74 (1H, br, H-4).
2.8.2.2. 11-Cyclohexyl-estra-5(10),9(11)-diene-3,17-dione
(21b). 21b was isolated as a white foam (0.09 g, 13%);
IR: 2929, 2952, 1739, 1719, 1669, 1593, 1450, 1367, 1226,
1061, 1013. HPLC: 96.9%; HRMS: calculated for C24H32O2
352.24023, found: 352.24039. 1H NMR {500 MHz, CDCl3
(TMS), δ (ppm)}: 0.86 (3H, s, H-18), 2.37 (1H, m, H-1ꢀ),
2.81 & 2.96 (2H, m, H-4).
for [M-18]+ C28H32O4 432.23006, found: 432.23015. H
1
NMR {500 MHz, CDCl3 (TMS), δ (ppm)}: 0.48 (3H, s,
H-18), 2.57 (3H, s, –CO–CH3), 3.87–4.06 (4H, m, ethylene-
dioxy), 4.35 (1H, d, 5-OH), 4.37 (1H, m, H-11), 7.34 (2H,
m, H-2ꢀ,6ꢀ), 7.86 (2H, m, H-3ꢀ,5ꢀ).
2.8.3. Procedure for preparation of the intermediates 22, 23
16a (1.0 g, 2 mmol) was added to a solution of pyridinium
p-toluenesulfonate (0.2 g, 0.8 mmol) in ethanol (30 ml).
The reaction mixture was stirred at 20 ◦C and and slowly
turned into a solution. The reaction was monitored by TLC
(n-hexane–ethyl acetate 1:1), and the intermediates were
2.12. 3-Ethylenedioxy-11-{4ꢀ-[1ꢀ,1ꢀ-(ethylenedioxy)-
ethyl]phenyl}-estra-5(10),9(11)-diene-17-one (25)
25 was prepared from 19a in 90% ethanol according to the
general procedure given for 22 by working up the reaction

744
E. Francsics-Czinege et al. / Steroids 68 (2003) 739–749
mixture after 10 min. TLC monitoring and column chro-
matography was carried out in n-hexane–ethyl acetate 1:1.
Yield: 0.26 g (27%); IR: 1739, 1604, 1253, 1199,
1111, 1037, 869, 836. HPLC: 97.5%; HRMS: calculated
for C30H36O5 476.25627, found: 476.25623. 1H NMR
{500 MHz, CDCl3 (TMS), δ (ppm)}: 1.01 (3H, s, H-18),
1.67 (3H, s, –CH3), 3.73–4.07 (8H, m, ethylenedioxy), 7.14
(2H, m, H-2ꢀ,6ꢀ), 7.37 (2H, m, H-3ꢀ,5ꢀ).
action mixture after 30 min. TLC monitoring and column
chromatography was carried out in n-hexane–ethyl acetate
1:1. Yield: 0.21 g (24%); IR: 1738, 1691, 1602, 1558,
1269, 1112, 1056, 836. HPLC: 97.8%; HRMS: calculated
for C28H32O4 432.23006, found: 432.23019. 1H NMR
{500 MHz, CDCl3 (TMS), δ (ppm)}: 1.02 (3H, s, H-18),
2.60 (3H, s, –CO–CH3), 3.80–3.96 (4H, m, ethylenedioxy),
7.28 (2H, m, H-2ꢀ,6ꢀ), 7.87 (2H, m, H-3ꢀ,5ꢀ).
2.13. 3-Ethylenedioxy-11-(4ꢀ-acetyl-phenyl)-estra-5(10),
9(11)-diene-17-one (26)
3. Results
26 was prepared from 19a in 90% ethanol according to
the general procedure given for 22 by working up the re-
A mixture of epoxides (5, 5␣,10␣/5␤,10␤ ratio 58:42)
was prepared from 3-ethylenedioxy-13␤-ethyl-gona-5(10),
O
O
O
H
H2O2, CCl3COCCl3,
H
HC
C CH2
O
H
6
H
O
O
H
CH2Cl2, 0 o
C
O
n-BuLi, THF, 0 o
C
O
4
O
5
OH
CH2
C
R
Mg Br
8a R=N(CH3)2
8b R= C(OCH2CH2O)CH3
O
O
C
H
H
O
H
O
7
Cu(I)Cl, THF, 10 oC
O
R
O
OH
CH2
C
O
O
C
H
H
H
5α,11β and 5β,11α
9a R=N(CH3)2
OH
O
9b R= C(OCH2CH2O)CH3
KHSO4/H2O or 2N HCl, MeOH
R
O
R
O
R
OH
OH
CH2OH
OH
CH2OH
C
CH2OH
C
C
C
C
C
H
H
H
H
H
H
O
10a R=N(CH3)2
R=COCH
12a R=N(CH3)2
12b R=COCH3
11 R=N(CH3)2
10b
3
Scheme 2.

E. Francsics-Czinege et al. / Steroids 68 (2003) 739–749
745
(Scheme 2). The stereoposition of the C(11) substituent
was verified, along with some characteristic chemical shift
data in NMR spectra, with NOE measurements: for an ␣
substituent the C(18) methyl group exhibited an NOE con-
nection with the H␤-11 proton, while for a ␤ substituent, it
showed NOE into the aromatic protons.
OMe
H
R
O
OH
H
R=
H
13a
13b R= OCH3
O
To the best of our knowledge, no 11␣-substituted com-
pound in the 13␤-series prepared by Teutsch’s method was
published, and the only example in the literature to demon-
strate the formation of a 5(10),9(11)-diene system was
described by Wiechert and coworkers [11] (13a-b)(Fig. 1).
13a-b were formed during the dehydration of pure
11␤-substituted 5␣-hydroxy intermediates. Their formation
was explained by steric factors due to ortho-substitution on
the phenyl ring.
After observing the formation of the unexpected prod-
ucts of type 12, the hydrolysis and dehydration were stud-
ied on model compounds to decide whether the by-products
could be formed only from one of the isomers (5␣,11␤ or
5␤,11␣) or from both of them with a different ratio. Model
Fig. 1. Known conjugated 5(10),9(11)-diene system.
9(11)-diene-17-one [10] with hexachloroacetone and 70%
hydrogen peroxide according to the procedure of Teutsch
et al. Nucleophilic addition of 2-[(prop-2-ynyl)oxy]-
tetrahydro-pyran (6) in the presence of n-butyl lithium gen-
erated the 17␣-propynyl side chain with a protected hydroxy
group (7). Conjugate addition of 8a or 8b Grignard-reagents
afforded the protected intermediates 9a and 9b, which were
hydrolyzed with KHSO4 in water (9a) or with 2N HCl (9b).
Products 10a and 10b were isolated by crystallization, while
11, 12a, and 12b were isolated by column chromatography
O
O
O
R
R
Cu(I)Cl, THF, 10 oC
H
H
HCl, EtOH
H
O
H
H
O
O
H
R
Mg Br
14
O
O
O
OH
O
O
O
CH
O
CH
C
16a R=
C
8b R=
17a R= 4-acetyl-phenyl
17b
R= cyclohexyl
16b R= cyclohexyl
15 R= cyclohexyl
O
O
R
H
H
15
8b or
O
H
O
H
O
O
O
CH
C
19a R=
Cu(I)Cl, THF, 10 o
C
18
O
OH
O
19b R= cyclohexyl
HCl, EtOH
O
O
R
HCl, EtOH
R
H
H
H
H
HCl, EtOH
O
O
20a R= 4-acetyl-phenyl
20b R= cyclohexyl
21a R= 4-acetyl-phenyl
21b R= cyclohexyl
Scheme 3.

746
E. Francsics-Czinege et al. / Steroids 68 (2003) 739–749
compounds with 11-aromatic and 11-cyclohexyl substitu-
tions were chosen to investigate the role of the conjugation
of the 5(10),9(11)-diene system with the 11-phenyl group
during formation of compounds type 12.
ilar behavior was observed in the case of compounds 16b
and 19b.
All protected isomers (16a, 16b, 19a, and 19b) were
deprotected and dehydrated according to the general prac-
tice by HCl in methanol. Hydrolysis of both 16a and 16b
afforded an only product, 17a and 17b, respectively, which
was expected from Teutsch’s procedure. However, both
11␣-isomers 19a and 19b resulted in two products following
hydrolysis. 20a and 21a were obtained from 19a at a ratio
of 35:65, while 19b yielded 20b and 21b at a ratio of 85:15.
20a and 21a could not be converted into each other by HCl
in methanol; both of them proved to represent a final stage
of hydrolysis. Both 11-cyclohexyl and 11-acetyl-phenyl
substituted model compounds led to the same type of de-
hydrated and deprotected products. The fact that 21b was
formed only at a yield of 15% versus 21a (65%) indicates
the possibility of spread delocalization of the ꢀ-electrons
in the 5(10),9(11)-double bonds; the 11-phenyl-group does
play an important role in the formation of the product (21a).
After observing the remarkable difference between the
behavior of 11␣ (19a and 19b) and 11␤ isomers (16a and
16b), we decided to take a closer look at the process of
deketalization and dehydration and to explore the pathways
leading to the formation of both the expected and unexpected
products. No intermediate could be isolated under general
hydrolytic circumstances (e.g. HCl in methanol); therefore,
milder conditions were necessary to investigate the chosen
11-(4-acetyl-phenyl)-derivatives 16a and 19a.
Model compounds 16a [5], 16b, 19a, and 19b were pre-
pared by conjugate addition of Grignard-reagents (8b, 15)
with the epoxides starting from both pure 5␣,10␣ (14) [12]
and 5␤,10␤ epoxide isomers (18) [13] (Scheme 3). 14 and
18 were isolated from a mixture of epoxides in a ratio of ␣/␤
67:33. The 5␣,10␣ is◦mer (14) was separated by repeated
crystallization from diisopropyl ether, while the 5␤,10␤ iso-
mer (18) was isolated by column chromatography, followed
by crystallization from diisopropyl ether. Taking advantage
of the selectivity of the copper(I) ion-catalyzed addition of
the Grignard-reagent at position 11 of the steroid skeleton,
protection of the 17-keto group was omitted.
As a matter of the rate of reactions, a marked differ-
ence was observed between the epoxides 14 and 18 in
the Grignard-reaction. 5␤,10␤-Epoxide (18) reacted with
the Grignard-reagents (8b and 15) much slower than the
5␣,10␣-epoxide (14). After carrying out the conjugate addi-
tion with the Grignard-reagents, all the protected intermedi-
ates (16a, 16b, 19a, and 19b) were isolated in crystal form.
The stereoposition of the C(11) substituent was verified,
along with some characteristic chemical shift data in the
NMR spectra, with NOE measurements: for an ␣ substituent,
the C(18) methyl group exhibited a NOE connection with
the H␤-11 proton, while for a ␤ substituent, it showed NOE
into the aromatic protons. Surprisingly, while 16a was found
to be stable, 19a was slowly decomposing, affording com-
pounds more apolar than 19a itself. Decomposition of 19a
was observed by TLC, and a remarkable difference be-
tween 16a and 19a was found by MS. In the EI mass spec-
trum of 16a, m/z 494 was detected, while for 19a, only
476 was found due to immediate H2O elimination. A sim-
Carrying out the hydrolysis under mild conditions, pyri-
dinium p-toluenesulfonate (PPTS), a weakly acidic salt was
applied in 90% ethanol and pure ethanol at 20 ◦C. The pro-
cess, monitored by TLC, was observed to be a series of
consecutive steps. All the intermediates were isolated when
they reached their maximum concentration in the reaction
mixture. Thus, 24, the only intermediate from 16a in 90%
O
O
O
C
C
O
H3C
O
H3C
O
H
H
H
H
O
PPTS
22
23
OEt
O
EtOH, 20 o
C
OEt
O
PPTS, EtOH, 20 oC, 24
or HCl ,EtOH, 20 oC, 5 min
17a
16a
O
C
PPTS
H3C
O
90 % EtOH, 20 oC
H
H
O
24
OH
O
Scheme 4.

E. Francsics-Czinege et al. / Steroids 68 (2003) 739–749
747
O
O
O
C
C
H3C
O
H3C
O
PPTS
EtOH, 20 oC
H
H
21a
19a
and in 90 % EtOH, 20 oC
H
H
O
O
O
O
25
26
EtOH, 20 oC, 5 hs
PPTS,
Scheme 5.
Table 1
ethanol, was isolated from the reaction mixture after 3.5 h
of reaction time, whereas PPTS in ethanol afforded two new
compounds 22 (after 20 min) and 23 (after 1 h) from 16a
(Scheme 4). After a long reaction time (24 h) the starting ma-
terial (16a) and the intermediates (22 and 23) disappeared,
forming 17a. In addition, both isolated 22 and 24 were con-
verted into 17a under the conditions described before.
In the case of 19a, the route leading to the formation of
major product 21a was mapped (Scheme 5) by TLC mon-
itoring and intermediate isolation. Both reaction conditions
provided the intermediates 25 (after 10 min in ethanol) and
26 (after 30 min in ethanol). Hydrolysis of 25 by PPTS
in ethanol yielded pure 21a, indicating that whenever the
5(10),9(11)-diene system conjugated with the 11-aryl group
formed no appearance of 3-keto-4,9-diene system could be
detected.
A remarkable distinction can be declared in the de-
hydration and deketalization of 5␣-hydroxy-11␤-, and
5␤-hydroxy-11␣-substituted 19-norsteroids indicated by the
different products. Moreover, 17a, the only product from
16a, and 21a, the major product from 19a, were observed
to be formed by different pathways. The selectivity of the
reaction conditions was proved by the differentiation of the
hydrolysis of the two ketals within the molecules. Despite
the selective conditions, elimination of the 5␣-hydroxy
group and hydrolysis of 3-ethylene ketal could not be
distinguished in 16a, showing that these steps proceeded
concomitantly. Conditions in abs. ethanol made it possible
to form ethyl ether from the relatively stable 5␣-hydroxy
group, rather than eliminate it.
Torsion angles of the hydrogens at positions 4 and 11 with respect to the
5␤OH–5C bond
Compounds
Torsion angles (◦)
5␤OH–5C–
4C–4␤Hax
5␤OH–5C–
4C–4␣Hekv
5␤OH–5C
–11C–11␤Hekv
19a
19b
39.8
45.1
76.0
71.0
8.1
0.2
rationalized by two crucial factors: possibility of conjuga-
tion of olefins (with the 3-oxo group or the 11-phenyl group)
and the orientation of the groups to be eliminated. Studies
of elimination in a variety of systems show that the orienta-
tion of groups to be eliminated determines reactivity in the
following order: anti-coplanar > syn-coplanar > nonplanar
[14]. It is obvious that in case of 11␤-derivatives (16a, 16b),
both the conjugation of the 4,9-diene system with the 3-oxo
group and the anticoplanar orientation of the 5␣-OH group
with the axial H at position 4, which is regarded to be the
most favorable for elimination, yielded only one type of
product 17a and 17b, respectively.
As the orientation of the eliminating groups was not ev-
ident in the 11␣-subsituted isomers, the three-dimensional
structures of the molecules 19a and 19b were modeled by
“HyperchemTM Release 4 for Windows Molecular Model-
ing System.” The geometry was optimized using the AM1
molecular approximation procedure and MM+ method,
and the torsion angles of the appropriate groups were
measured (Table 1). To confirm our explanation for the
11␤-derivatives, 16a and 16b were also modeled (Table 2).
Figs. 2 and 3 show the three-dimensional structures of 16a
and 19a to demonstrate the difference in geometry between
the 11␣- and 11␤-isomers.
However, under the same conditions 19a was compelled
to eliminate the 5␤-hydroxy group, mainly in vinylogous
syn-elimination with the periplanar 11␤-hydrogen first.
This elimination preceded not only the hydrolysis of the
3-ethylene-ketal, but of the one on the 11␤-(4-acetyl-phenyl)
substituent, which proved to be the fastest step in the for-
mation 17a.
Table 2
Torsion angles of the hydrogens at positions 4 and 11 with respect to the
5␣OH–5C bond
Compounds
Torsion angles
5␣OH–5C–
4C–4␤Hax (◦)
5␣OH–5C–
4C–4␣Hekv
5␣OH–5C–
11C–11␣Hekv
4. Discussion
16a
16b
5.1
5.3
58.2
58.6
39.0
71.8
Relying upon the above results, we assume that the out-
come of the elimination and deprotection reactions can be

748
E. Francsics-Czinege et al. / Steroids 68 (2003) 739–749
Fig. 2. The three-dimensional structure of 19a optimized by HyperchemTM Release 4.
Fig. 3. The three-dimensional structure of 16a optimized by HyperchemTM Release 4.

E. Francsics-Czinege et al. / Steroids 68 (2003) 739–749
749
On the basis of the torsion angles measured in the
three-dimensional structures of 19a and 19b, it was obvi-
ous that in these molecules the orientation of the 11␤H and
5␤OH groups is syn-coplanar; consequently, the overlap of
the orbitals was favorable for syn-elimination and was more
favorable, than for any of the hydrogens at position 4, which
were nonplanar. In the case of 19a, the forming 5(10),9(11)-
diene-system was stabilized by conjugation with the 11-
(4-acetyl-phenyl)-group, resulting in a thermodynamically
stable compound 21a. Despite the favorable orientation of
11␤H and 5␤OH in 19b for syn-elimination, the product
containing the 5(10),9(11)-diene-system was only a minor
component, as there was no possibility for conjugation with
the 11-substitution. The major product 20b, stabilized by
the 3-keto-4,9-diene system was more stable thermodynam-
ically.
[4] Rao PN, Cessac JW, Blye RP, Kim HK. 11␤-Substituted 13␤-ethyl
gonane derivatives exhibit reversal of antiprogestational activity.
Steroids 1998;63:50–7.
[5] Rao PN, Wang Z, Cessac JW, Rosenberg RS, Jenkins DJA, Diamandis
EP. New 11␤-aryl substituted steroids exhibit both progestational and
antiprogestational activity. Steroids 1998;63:523–30.
[6] Teutsch G, Bélanger A. Regio and stereospecific synthesis of 11␤-
substituted 19 norsteroids. Tetrahedron Lett 1979;22:2051–4.
[7] Jones RG, Mann MJ. New methods of synthesis of ␤-aminoe-
thylpyrazoles. J Am Chem Soc 1953;75:4048–52.
[8] Neville RG. Synthesis of 4-acetylphenylmethylsilanes using 2-(4-
bromophenyl)-2-methyl-1,3-dioxolane. J Org Chem 1959;24:111–2.
[9] Miyashita M, Yoshikoshi A, Grisco PA. Pyridinium p-toluenesul-
fonate. A mild and efficient catalyst for tetrahydropyranylation of
alcohols. J Org Chem 1977;42:3772–4.
[10] Teichmüller G, Müller G, Krahmer S, Streibhardt G, Wentzke
M. Preparation of 17 alpha-cyanomethyl-17 beta-hydroxy steroids.
(Jenapharm, Germany) Ger. (East) DD 296495 A5 1991125
6pp (German). APPLICATION: DD 90-342604 19900709. CA
116:129375.
[11] Cleve A, Ottow E, Neef G, Wiechert R. Synthesis of oxygen-bridged
antigestagens. Tetrahedron 1993;49:2217–26.
References
[12] Faraj H, Claire M, Rondot A, Aumelas A, Auzou G. Synthesis
of new steroidal 11␤-substituted spirolactones. J Chem Soc Perkin
Trans 1 1990;11:3045–8.
[13] Zhang TY, Lee YW, Fang QC, Xiao R, Ito Y. Preliminary applications
of cross-axis syncronous flow-through coil planet centrifuge for large-
scale preparative counter-current chromatography. J Chromatogr
1988;454:185–93.
[1] Bélanger A, Philibert D, Teutsch G. Regio and stereospecific
synthesis of 11␤-substituted 19-norsteroids. Steroids 1981;37:361–
82.
[2] Teutsch G, Philibert D. History and perspectives of antiprogestins
from the chemist’s point of view. Human Reprod 1994;9:12–
31.
[14] Kirk DN, Harsthorn MP. Steroid reaction mechanism In: Eaborn
C, editor. Reaction mechanism in organic chemistry. A series
of monographs (monograph 7). Amsterdam: Chapman NB; 1968.
p. 101–27.
[3] Cook CE, Lee YW, Wani MC, Fail PA, Petrow V. Effects of D-
ring substituents on antiprogestational (antagonist) and progestational
(agonist) activity of 11␤-aryl steroids. Human Reprod 1994;9:
32–9.