Neighboring group participation: Part 15. Stereoselective synthesis of some steroidal tetrahydrooxazin-2-ones, as novel presumed inhibitors of human 5α-reductase

Article abstract of DOI:10.1016/j.steroids.2004.04.003

During the alkaline methanolysis of 3β-acetoxy-21-chloromethyl-pregn- 5-ene-20β-N-phenylurethane, and its p-substituted phenyl derivatives, cyclization occurs, in the course of which 17β-[3-(N-phenyl) tetrahydrooxazin-2-on-6-yl]androst-5-en-3β-ol and its p-substituted phenyl derivatives are formed. The cyclization takes place with (N^--6) neighboring group participation. Oppenauer oxidation of the 3β-hydroxy-exo- heterocyclic steroids yielded the corresponding Δ⁴-3- ketosteroids. The structures of the new compounds were proved by IR, ¹H and ¹³C NMR spectroscopy, using up-to-date measuring techniques such as 2D-COSY, HMQC, and HMBC. The inhibitory effects (CI ₅₀) of the Δ⁴-3-ketosteroids on 5α-reductase were studied.

Full text of DOI:10.1016/j.steroids.2004.04.003

Steroids 69 (2004) 451–460
Neighboring group participation
Part 15. Stereoselective synthesis of some steroidal
tetrahydrooxazin-2-ones, as novel presumed
inhibitors of human 5␣-reductaseଝ
János Wölflinga, László Hacklera, Erzsébet Mernyáka, Gyula Schneidera,∗, István Tóthb,
Mihály Szécsib, János Juleszb, Pál Sohárc,d, Antal Csámpaic
a
Department of Organic Chemistry, University of Szeged, Dóm tér 8, H-6720 Szeged, Hungary
b
Endocrine Unit and Research Laboratory, University of Szeged, Korányi fasor 8, H-6720 Szeged, Hungary
c
Department of General and Inorganic Chemistry, Eötvös Lóránt University, P.O. Box 32, H-1518 Budapest, Hungary
d
Research Group of Structural Chemistry and Spectroscopy, Hung. Acad. Sci., P.O. Box 32, H-1518 Budapest, Hungary
Received 10 February 2004; received in revised form 2 April 2004; accepted 8 April 2004
Available online 17 June 2004
Abstract
During the alkaline methanolysis of 3␤-acetoxy-21-chloromethyl-pregn-5-ene-20␤-N-phenylurethane, and its p-substituted phenyl
derivatives, cyclization occurs, in the course of which 17␤-[3-(N-phenyl)tetrahydrooxazin-2-on-6-yl]androst-5-en-3␤-ol and its p-sub-
stituted phenyl derivatives are formed. The cyclization takes place with (N−-6) neighboring group participation. Oppenauer oxidation of
the 3␤-hydroxy-exo-heterocyclic steroids yielded the corresponding ꢀ4-3-ketosteroids. The structures of the new compounds were proved
1
by IR, H and 13C NMR spectroscopy, using up-to-date measuring techniques such as 2D-COSY, HMQC, and HMBC. The inhibitory
effects (CI50) of the ꢀ4-3-ketosteroids on 5␣-reductase were studied.
© 2004 Elsevier Inc. All rights reserved.
Keywords: Heterocyclic steroids; Neighboring group participation; 5␣-Reductase; In vitro inhibition; Structure determination by IR and NMR
1. Introduction
analog were reported recently; these display a high inhibitory
activity against P45017␣. It has been suggested that this
activity is related to the presence of the heterocyclic moiety
in ring D [4]. Guarna et al. have reported on the synthe-
sis of (+)-17-(3-pyridyl)-(5␤)-10-azaestra-1,16-dien-3-one,
which may be a fundamental 5␣-reductase inhibitor [5].
Since 5␣-reductase converts testosterone to the more po-
tent androgen dihydrotestosterone in prostatic tissues, the
dual inhibition of 5␣-reductase and P45017␣ might be par-
ticularly beneficial in the treatment of prostate cancer [6].
Therefore, it was hypothesized that anchoring a 3ꢀ-pyridyl
moiety on the D ring of 19-nor-10-azasteroid would pro-
duce novel steroid derivatives, as presumed inhibitors of
both 5␣-reductase and P45017␣ [7].
Prostatic cancer is the second leading cause of cancer-
related mortality worldwide. Considerable attention has
been devoted to the development of drugs that are active
against prostate cancer by intervening in the processes of
tumor growth, and especially in the steroid metabolism
pathways of testosterone production [2]. Androgen depri-
vation by the inhibition of human cytochrome P45017␣
(17␣-hydroxylase/C17,20-lyase), the enzyme responsible
for the conversion of C21 steroids to C19 androgens, has
been proposed as a potential approach for the therapeu-
tic treatment of this disease, which is androgen-dependent
in the majority of cases [3]. The syntheses of abiraterone
[17-(3-pyridyl)androsta-5,16-dien-3␤-ol] and its 4-en-3-one
We set out to synthesize a novel series of steroidal
tetrahydrooxazin-2-ones containing heterocycles involving
O and N heteroatoms at position 17␤ of androst-5-en-3␤-ol
and position 17␤ of androst-4-en-3-one, respectively.
The present work was stimulated by the importance
ଝ
For a preceding paper in this series, see reference [1].
Corresponding author. Tel.: +36-62-544276; fax: +36-62-544200.
∗
E-mail address: schneider@chem.u-szeged.hu (G. Schneider).
0039-128X/$ – see front matter © 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.steroids.2004.04.003

452
J. Wölfling et al. / Steroids 69 (2004) 451–460
1
of pentacyclic steroids offering novel binding sites and
binding distances to the receptors by introducing het-
eroatoms and various functionalities into the exo-hete-
rocycles.
The characteristic IR band frequencies and the H and
13C NMR data on the compounds investigated (3d, 4d, 5a,
b, 6a–h, 7a–h, and 8a–h) are listed in Tables 1–3.
2.2. 21-Hydroxymethylpregn-5-ene-3β,20-diol (3a)
The discovery of two different genes encoding for two
5␣-reductase isoenzymes, types 1 and 2 [8], with differ-
ent biochemical characteristics and tissue distributions [9],
added a further impulse to this research. In human, the type
1 isoenzyme is present mainly in the nongenital skin and
in the liver whereas the type 2 is active in the prostate,
genital skin, epididymis, seminal vesicle, and also in the
liver. Cancerous and hyperplastic prostate are abundant in
the type 2 isoenzyme, therefore tissue specimens of these
tumors are suitable enzyme sources for the in vitro study of
the 5␣-reductase type 2 inhibitors. In vitro experiments on
the newly synthesized steroid compounds 8a–g were carried
out with homogenates of human prostates.
21-Hydroxymethylenepregn-5-en-3␤-ol-20-one (2) (3.44
g, 10 mmol), prepared by the procedure of Hirsch and Fu-
jimoto [10] (mp 209–211 ◦C), was suspended in ethanol
(50 ml) and cooled in an ice-bath, and KBH4 (1.62 g,
0.03 mol) was added in small portions. The mixture was
allowed to stand for 6 h, and was then neutralized with
dilute HCl. The resulting solution was poured onto ice
(300 g), and the precipitate that separated out was fil-
tered off, dried over P2O5, and crystallized from ace-
tone/water to give 3a (3.05 g, 87%), mp 184–186 ◦C, Rf
= 0.30 (ss A); [␣]D − 26 (c 1 in methanol) ([10] mp
183–185 ◦C).
2.3. 3β-Acetoxy-21-hydroxymethylpregn-5-en-20β-ol (3d)
and 3β-acetoxy-21-hydroxy methylpregn-5-en-20α-ol (4d)
2. Experimental
2.1. General
3␤-Acetoxypregn-5-ene-21-hydroxymethyl-20-hydroxy
cyclic acetonide (3b) (4.30 g, 10 mmol), prepared by the
procedure of Hirsch and Fujimoto [10] (mp 154–166 ◦C),
was dissolved in methanol (60 ml) and p-toluenesulfonic
acid (0.5 g) in water (3 ml) was added to the solution.
The reaction mixture was allowed to stand for 6 h and
was then poured onto water (500 ml). The precipitate
that separated out was filtered off, dissolved in chloro-
form, and subjected to chromatographic separation on
silica gel in ethyl acetate/chloroform (5:95), yielding
pure 3d (3.1 g, 79%), mp 165–168 ◦C, Rf = 0.35 (ss C);
Melting points (mp) were determined on a Kofler block
and are uncorrected. Specific rotations were measured in
CHCl3 (c 1) at 20 ◦C with a POLAMAT-A (Zeiss-Jena)
polarimeter and are given in units of 10−1 degree cm2 g−1
.
Elemental analysis was carried out in the analytical labora-
tory of the University of Szeged with a Perkin-Elmer model
2400 CHN analyzer. The reactions were monitored by TLC
on Kieselgel-G (Merck Si 254 F) layers (0.25 mm thick);
solvent systems (ss) (A): acetone/benzene/light petroleum
(30:35:35 v/v), (B): ethyl acetate/chloroform (10:90 v/v),
(C): ethyl acetate/chloroform (5:95 v/v), (D): ethyl ac-
etate/chloroform (1:99), and (E): dichloromethane. The
spots were detected by spraying with 5% phosphomolyb-
dic acid in 50% aqueous phosphoric acid. The Rf values
were determined for the spots observed by illumination
at 254 and 365 nm. Flash chromatography: silica gel 60,
40–63 ␮m. All solvents were distilled prior to use. The
IR spectra were run in KBr disks on a Bruker IFS-113v
vacuum optic FT-spectrometer with an Aspect 2000 com-
puter (3d and 6–8a, b, d, f, h) or on a Bruker IFS-55
FT-spectrometer controlled by Opus 3.0 software (4d, 5a, b
and 6–8c, e, g). The 1H and 13C NMR spectra were recorded
in CDCl3 or in (CD3)2SO solution in 5 or 10 mm tubes at
room temperature, on a Bruker WM-250 and/or WP-80-SY
FT-spectrometer controlled by an Aspect 2000 computer at
250.13 MHz (1H) and 62.89 or 20.14 MHz (13C), and on a
Bruker DRX 500 spectrometer at 500.13 (1H) and 125.76
(13C) MHz, with the deuterium signal of the solvent as the
lock and TMS as internal standard. DEPT spectra were run
in a standard manner, using only the Θ = 135◦ pulse to
separate CH/CH3 and CH2 lines phased “up” and “down,”
respectively. The HMQC and HMBC spectra were obtained
by using the standard Bruker pulse programs.
20
[␣]D − 53 (c 1 in chloroform) (found: C, 73.76; H,
9.95. C24H38O4 requires C, 73.81; H, 9.81%); contin-
ued elution resulted in a mixture of 3d and 4d (500 mg,
13%).
Continued elution furnished 4d (170 mg, 4%), mp
20
183–186 ◦C, Rf = 0.25 (ss B); [␣]D − 45 (c 1 in chlo-
roform) (found: C, 73.58; H, 9.65. C24H38O4 requires C,
73.81; H, 9.81%).
2.4. 3β-Acetoxy-21-p-toluenesulfonyloxymethylpregn-5-
ene-20β-ol (5a)
Compound 3d (3.9 g, 10 mmol) was dissolved in pyridine
(30 ml), and a solution of p-toluenesulfonyl chloride (2.1 g,
11 mmol) dissolved in pyridine (10 ml) was added during
cooling in ice. The reaction mixture was allowed to stand at
room temperature for 20 h, and was then saturated with wa-
ter. The precipitate was filtered off, dissolved in chloroform
and chromatographed on silica gel with a mixture of ethyl
acetate/chloroform (2.5:75.5) to obtain 5a (4.8 g, 88%), mp
20
176–178 ◦C, Rf = 0.45 (ss C); [␣]D − 20 (c 1 in chlo-
roform) (found C, 68.42; H, 8.02. C31H44O6S requires C,
68.35; H, 8.14%).

Table 1
1H NMR dataa and characteristic IR frequenciesb of compounds 3d, 4d, 5a, b, 6a–h, 7a–h, and 8a–h
c
d
c
c
g,h
=
Compound
CH3
CH3
(Ar)
CH3
CH3
H-3e
m (1H)
H-6/4f
(1H)
d
OCHg
(1H)
m
XCH2
OH/NH
br (1H)
ArH-2ꢀ,6ꢀ
ArH-3ꢀ,5ꢀ
␯C O
␯NCOO band
␥CArH
band
(OAc)
Pos. 18
Pos. 19
m (2H)
∼d (2H)i
∼d (2H)i
bandj
3d
4d
5a
5b
6a
6b
6c
6d
6e
6f
6g
6h
7a
7b
7c
7d
7e
7f
7g
7h
8a
8b
8c
8d
8e
8f
2.03
2.03
2.02
2.02
2.03
2.03
2.04
2.03
2.03
2.04
2.04
2.04
−
−
0.79
0.70
0.73
0.78
0.70
0.74
0.75
0.74
0.74
0.74
0.74
0.74
0.78
0.78
0.85
0.84
0.76
0.84
0.84
0.84
0.88
0.87
0.86
0.87
0.80
0.87
0.86
0.87
1.03
1.02
1.01
1.02
0.98
0.99
1.00
0.99
1.02
1.00
0.99
0.99
0.97
0.97
1.03
1.03
1.01
1.03
1.03
1.03
1.19
1.19
1.19
1.20
1.20
1.20
1.19
1.20
4.62
4.60
4.59
4.59
4.62
4.60
4.61
4.62
4.60
4.60
4.61
4.65
∼3.25m
∼3.25m
3.52
∼3.6n
3.51
∼3.65n
3.53
∼3.55
−
5.40
5.38
5.36
5.36
5.36
5.3
∼3.85
3.85
3.63
3.75
4.95
5.00
5.02
5.00
5.10
5.00
5.02
5.05
4.38
4.35
4.36
4.35
4.32
∼4.4
4.36
4.36
4.36
4.35
4.34
4.35
4.34
∼4.4
4.36
4.38
∼3.85
3.85
4.14
3.70
4.12
3.56
3.57
3.57
3.56
3.56
3.56
3.56
3.55
∼2.65h
−
−
1733
1732
1713
1717
1731
1740
1721
1740
1731
1735
1734
1740
−
−
−
−
∼2.65
−
−
−
−
2.44
−
7.78
−
7.34
−
903
−
−
−
2.36
2.30
1.22
3.79
1.39
6.69
6.73
6.73
6.62
∼6.6
6.95
6.91
∼6.8
4.56
4.50
7.74
7.10
7.33
∼7.35
∼7.3
7.00
7.26
7.2–7.5
7.22k
∼7.20
∼7.22
7.25
7.19
7.21
7.26
∼7.05
∼7.3
∼7.18
∼7.20
7.21
7.19
7.20
7.25
∼7.1
7.05–7.4k
7.30
1731
1723
1721
1722
1731
1716
1714
1724
1667
1676
1672
1680
1686
1692
1702
1673
1695
1680
1686
1683
1686
1700
1682
1692
813l
817
832
829
830
833
828
822
759l
823
840
832
830
827
826
837
754l
816
840
850
840
824
845
850
5.37
5.38
5.37
5.36
5.36
5.40
5.27
5.27
5.35
5.35
5.35
5.40
5.36
5.35
5.72
5.72
5.71
5.72
5.74
5.73
5.72
5.72
∼7.14
6.85
6.83
∼7.4
∼7.38
7.2–7.5
∼7.35k
∼7.20
∼7.22
6.90
6.87
7.49
7.35
∼7.35
∼7.3
∼7.18
∼7.20
6.90
6.88
7.49
7.33
∼7.3
−
−
2.29
1.24
3.80
1.40
−
3.55
−
3.62
−
∼3.6n
−
−
3.57
−
−
∼3.65n
3.62
−
−
−
∼1.6
−
−
−
∼3.75
∼3.7
∼3.65
3.62
−
−
−
−
−
−
−
−
−
−
−
1675
1680
1670
1683
1686
1670
1682
1670
−
2.33
1.22
3.80
1.40
−
−
−
−
−
−
∼3.72
−
−
3.58
−
−
∼3.7
3.62
∼3.65
8g
8h
−
−
−
−
−
−
a Measuring frequency: 500 (4d, 5a, b, and 6–8c, e, g) or 250 MHz (3d and 6–8a, b, d, f, h). Solvent: CDCl3 or DMSO-d6 (for 7a, b). Chemical shifts in ppm (δTMS = 0 ppm), coupling constants in
Hz. Assignments were supported by 2D-HMQC (for 4d, 5a, 6c, 7g, and 8c) and for 4d also by 2D-COSY measurements.
b In KBr discs (cm−1). Further bands, ␯OH (doubled for 7e, h): 3380–3500 (3d, 4d, 5a, b, and 7a–h), ␯NH: 3315–3430 (6a–h, doubled for 6d, e), ␯SO2: 1358, 1190, 1178 (5a and 6a).
c s (3H).
d s (3H), OMe group (6–8d), (Ar)Et group (6–8c, e): t (J: 7.0), CH2 (ArEt): 2.62 (6–8c), OCH2 (ArEt): 4.02 (7e and 8e).
e For 5a, b and 6a–h tt with coalesced lines.
f Pos. 6 (3–7), Pos. 4 (8),∼d, J: 4.5
0.6 (5, 6), 5.1
0.1 (7), ∼1 (8).
g In the side-chain or oxazinone ring, X: O (3d, 4d, and 5a, 6a), Cl (5b and 6b–h), or N (7 and 8).
h Two m’s (2 × 1H) with the second signal at ∼2.95 (2H, 3d), 4.25 (5a), 3.72
0.03 (7a, b, c, g, and 8c, g), 3.64
0.02 (7e and 8e).
i Rudimentaly AAꢀBBꢀ-type spectrum: J: 8.3 (5a and 6a–c), 8.8 (5d, e, g, 7d–g, and 8d–g), ∼s, broad signals with coalesced fine structure (due to F,H-couplings for 6–8h): 2 × m (2 × 2H, 6–8h),
m (4H) for 6h, 7b, c, and 8b, c, m (5H) for 8a.
j Ester (3d, 4d, 5a, b, and 6a–h) or ketone group (8a–h).
k Intensity: 7/1/4H.
l p-Disubstituted benzene ring, 748 (phenyl), ␥CArCAr (phenyl): 698 (6a, 7a, and 8a).
m Hidden by the water signal of the solvent.
n Overlapping signals.

454
J. Wölfling et al. / Steroids 69 (2004) 451–460
Table 2
13C NMR chemical shiftsa of compounds 3d, 4d, 5a, b, 6a–h, 7a–h, and 8a–hb,c
Compound C-1
C-2
C-3
C-4
C-5
C-6
C-7
C-8
C-9
C-10 C-11 C-12 C-13 C-14 C-15 C-16 C-17 C-18 C-19
3d
4d
5a
5b
6a
6b
6c
6d
6e
6f
6g
6h
7a
7b
7c
7d
7e
7f
7g
7h
8a
8b
8c
8d
8e
8f
38.0d 37.1
38.4d 37.4
37.4 36.3
38.5 37.4
38.2 37.1
38.1d 37.0
38.4d 37.4
38.1d 37.1
38.5d 37.4
38.0d 37.1
38.5d 37.4
38.2d 37.0
38.7 33.1e 71.7
38.6 33.0e 71.7
37.6 32.1d 72.2
37.4 32.1d 71.9
37.7 32.2d 72.0
38.5 33.0e 71.6
74.0d 39.8 139.8 122.4 31.9 31.8 50.1 36.6 21.0 38.1d 42.4 56.2d 25.4 27.8 56.5d 12.3 19.3
74.3
74.3
74.3
74.0
74.0
74.4
39.3 140.1 122.9 32.2 31.9 50.4 37.0 21.1 38.5d 42.1 56.9d 25.4 28.2 57.1d 13.1 19.7
38.5 140.1 122.8 32.3 32.1 50.4 37.0 21.3 40.3 42.8 56.9 25.6 28.1 56.4 12.8 19.7
40.0d 140.1 122.8 32.3 32.1 50.4 37.0 21.3 40.3d 42.9 57.0 25.7 28.1 56.5 12.8 19.7
39.1 140.0 122.3 31.9e 31.9e 50.3 36.7 21.0 33.6 42.4 56.1 25.4 27.9 53.4 12.3 19.2
39.1d 139.9 122.3 31.9 31.8 50.2 36.6 21.0 38.2d 42.4 56.1 25.5 27.8 53.6 12.4 19.3
38.5d 139.9 122.8 32.3 32.1 50.5 37.0 21.4 39.5 42.8 56.4 25.9 28.2 54.0 12.8 19.7
74.0d 39.1 139.9 122.3 31.8e 31.8e 50.2 36.6 21.0 38.2d 42.5 56.1 25.5 27.8 53.7 12.3 19.3
74.3e 39.1 140.1 120.8 32.2 31.9 50.4 37.0 21.1 38.7d 42.2 56.8 25.1 28.1 54.6 12.9 19.7
74.0d 39.2 139.9 122.3 31.9 31.8 50.2 37.0 21.1 38.7d 42.2 56.8 25.1 28.1 54.6 12.9 19.7
74.4
74.0
39.5d 140.2 120.3 32.2 32.1 50.5 37.0 21.4 38.3d 42.8 56.4 25.8 28.2 53.9 12.8 19.7
39.1d 139.9 122.3 31.83 31.75 50.2 36.7 210 37.9d 42.4 56.1 25.5 27.8 53.5 12.4 19.3
43.9 143.2 121.8 33.1e 33.1e 51.6 37.8 22.0 40.4 43.7 57.4 25.7d 28.1 54.9 13.7 20.8
43.8 143.1 121.6 33.0e 33.0e 51.5 37.8 21.9 40.3 43.6 57.3 25.5d 28.0 54.9 13.6 20.6
42.7 143.1 121.7 32.4d 32.2 50.6 37.0 21.2 39.4 43.0 56.7 25.1d 27.3 54.0 12.7 19.8
42.4 143.3 121.3 31.9d 32.0 50.5 36.7 20.9 39.2 42.7 56.5 24.8d 27.1 53.9 12.3 19.5
42.6 141.2 121.8 32.0d 31.9 50.5 36.9 21.2 39.3 42.2 56.8 26.0d 28.1 54.9 13.0 19.2
43.8 143.1 121.5 33.0e 33.0e 51.5 37.7 21.9 40.2 43.5 57.2 25.5d 27.9 54.8 13.5 20.5
42.7 141.5 121.7 32.4 32.2 50.6 37.0 21.2 39.4 43.0 56.7 24.9d 27.2 54.1 12.7 19.8
42.6 141.4 121.3 32.0d 32.1d 50.6 36.8 21.0 39.3 42.8 56.6 24.9d 27.1 54.0 12.3 19.5
37.6 32.1
37.5 32.2d 71.8
72.2
35.6d 33.9d 199.1 123.8 170.9 32.8d 32.2d 35.7d 53.7 38.7 20.8 39.0 42.7 55.5 24.6d 26.9d 54.0 12.3 17.4
35.7d 34.0d 199.5 123.8 171.3 32.9d 32.1d 35.6d 53.7 38.7 20.8 38.9 42.7 55.5 24.7d 26.9d 53.9 12.3 17.4
36.1 34.4 200.0 124.2 171.7 33.3 32.6 35.9 54.3 39.1 21.2 39.3 43.0 55.8 24.8d 27.2d 54.0 12.7 17.8
35.7d 33.9 199.4 123.8 171.3 32.8 32.1 35.5d 53.6d 38.6 20.8 38.9 42.6 55.4e 24.6d 26.9d 53.9d 12.3 17.4
36.1 34.4 199.8 124.3 171.4 33.2 32.3 35.6 54.1d 39.0 21.1 39.1 42.3 55.9 24.5d 28.2d 54.9d 13.2 17.8
35.6 33.8 199.1 123.7 171.0 32.7d 32.0d 35.4 53.6d 38.5 20.7 38.8 42.5 55.3 24.4d 26.7d 53.8d 12.2 17.3
36.1 34.4 200.0 124.2 171.7 33.3d 32.5d 35.9 54.0d 39.1 21.2 39.3 43.0 55.8 24.8d 27.2d 54.3d 12.8 17.8
35.8 34.0 199.4 123.9 171.1 32.9d 32.2d 35.6 53.7d 38.7 20.8 39.0 42.7 55.5 24.4d 26.9d 54.0d 12.4 17.5
8g
8h
a In ppm (δTMS = 0 ppm) at 126.7, 62.9 (3d, 6b, f, h, 7b, d, f, h, and 8b, d, f, h) or 20.1 MHz (6a, d, 7a, and 8a). Solvent: CDCl3 or DMSO-d6
(for 7a, b), respectively.
b Assignments were supported by DEPT (except for 3d, 6a, 7b, f, h, and 8a, b, d), and for 4d, 5a, 6c, 7g, and 8c also by HMQC and HMBC
measurements.
c Further lines: see Table 3 and the footnote to it.
d Interchangeable assignments.
e Overlapping lines.
dichloromethane fraction was washed thoroughly with wa-
ter, dried and evaporated to dryness. The residue was sub-
2.5. 3β-Acetoxy-21-chloromethylpregn-5-ene-20β-ol (5b)
jected to chromatographic separation with dichloromethane.
Compound 5b (818 mg; 2 mmol) was dissolved in ben-
Compound 3d (3.9 g, 10 mmol) was suspended in CCl4
(40 ml) and Ph3P (2.88 g, 3.1 mmol) was added. After an 6-h
zene (25 ml) and triethylamine (0.3 ml) and a solution of
reflux, the solvent was evaporated off. The resulting crude
phosgene in benzene (5 ml; 10%) were added. The reaction
product was chromatographed on silica gel with a mixture
mixture was allowed to stand at room temperature for 24 h.
of ethyl acetate/chloroform (2.5:97.5) to obtain 5b (2.9 g,
Diethyl ether (30 ml) was then added, and the triethylamine
HCl that separated out was filtered off and the filtrate was
evaporated in vacuo. The residue was dissolved in benzene
(30 ml), and triethylamine (0.3 ml) and the corresponding
substituted aniline (2.2 mmol) were added. The mixture was
allowed to stand for 6 h, and was then diluted with water. The
precipitate that separated out was filtered off and dried over
P2O5. The resulting crude product was chromatographed on
silica gel with dichloromethane.
74%), mp 183–184 ◦C, Rf = 0.50 (ss C); [␣]D − 31 (c 1 in
20
chloroform) (found C, 70.56; H, 8.98. C24H37O3Cl requires
C, 70.48; H, 9.12%).
2.6. 3β-Acetoxy-21-p-toluenesulfonyloxymethylpregn-5-
ene-20β-N-phenylurethane (6a) and 3β-acetoxy-21-
chloromethylpregn-5-ene-20β-N-phenylurethane (6b)
2.6.1. General procedure
Compound 5a (1.09 g, 2 mmol) or 5b (818 mg; 2 mmol)
was dissolved in dichloromethane (25 ml) and phenyl iso-
cyanate (960 mg, 8 mmol) and triethylamine (0.1 ml) were
added. The reaction mixture was refluxed for 6 h, and
was then poured into 10% NaHCO3 solution (50 ml). The
2.6.2. 3β-Acetoxy-21-p-toluenesulfonyloxymethylpregn-
5-ene-20β-N-phenylurethane (6a)
6a (1.18 g; 85%), mp 166–167 ◦C, Rf = 0.6 (ss D);
20
[␣]D − 51 (c 1 in chloroform) (found C, 68.82; H, 7.55.
C38H49O7SN requires C, 68.75; H, 7.44%).

J. Wölfling et al. / Steroids 69 (2004) 451–460
455
Table 3
13C NMR chemical shiftsa of compounds 3d, 4d, 5a, b, 6a–h, 7a–h, and 8a–hb,c
=
Compound
CH3 (OAc)
CH3 (Ar)
Side chain or oxazinone ring
Aryl group
C-1ꢀ
NCOO
C O (OAc)
CH(OH)
CCH2C
XCH2
C-2ꢀ,6ꢀ
C-3ꢀ,5ꢀ
C-4ꢀ
3d
4d
5a
5b
6a
6b
6c
6d
6e
6f
6g
6h
7a
7b
7c
7d
7e
7f
7g
7h
8a
8b
8c
8d
8e
8f
21.3
21.8
21.8
21.8
21.3d
21.4
21.8
21.3
21.8
21.4
21.9
21.4
−
−
74.3d
74.6
70.8
71.9
73.6
74.5
74.8
74.5
74.3d
74.8
75.2
74.8d
80.4
80.2
79.6
79.3
80.7
80.4
79.9
79.5
79.1
79.1
79.5
79.1
80.2
79.2
79.8
79.3
24.6
24.5
24.9
24.9
24.3
24.4
24.7
24.3
24.4
24.4
24.4
24.3
25.6d
25.6d
24.9d
24.8d
24.6d
25.4d
25.1d
24.6d
26.9
26.9
27.2
26.9
28.2
26.7
27.2
26.9
61.4
62.3
68.3
42.3
67.1
40.7
41.1
40.6
41.3
0.6
41.0
40.6
48.7
48.7
48.2
48.2
48.8
48.3
48.0
48.0
47.6
47.8
48.2
48.1
48.8
47.3
48.0
48.0
−
−
−
−
−
−
170.6
171.0
170.9
170.6
170.4
170.5
170.6
170.4
170.6
170.6
171.2
170.7
−
−
−
−
−
−
22.0
−
133.6
−
128.3
−
130.3
−
145.2
−
−
−
21.5d
20.7
16.2
55.0
15.3
−
133.5e
133.0
136.0
131.3
131.4
134.1
137.1
137.4
145.2
142.6
141.1d
136.4
136.2
144.3
142.0
139.2
143.3
140.6
141.1
136.1
136.1
142.1
142.0
139.1f
129.0e
119.0
119.2
121.0
122.8
∼120.7
122.7
120.4
127.4
127.1
126.1
127.2
127.6
129.2
127.4
127.6f
125.7
125.7
126.1
127.1
127.6
127.1
127.4
127.6f
129.8e
129.5
128.8
114.4
115.3
115.6f
129.4
131.9
130.3
130.6
129.0
114.6
115.4
132.9
129.6
116.0f
129.0
129.8
129.0
114.4
115.5
131.9
129.6
116.0f
144.6e
135.5
140.2
156.2
155.7
159.1f
128.8
115.9
127.5
136.7
141.6d
158.3
157.9
119.7
132.4
161.0f
126.4
136.4
143.1
158.1
158.0
119.5
132.5
161.0f
152.9
153.2
153.5
153.5
154.1
153.3
153.4
153.0
153.2
153.6
153.1
152.8
153.8
152.9
152.8
152.6
152.3
152.6
153.0
152.8
153.7
152.0
152.7
152.5
−
−
−
−
22.0
15.8
55.6
15.2
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
21.0
15.8
55.4g
15.2
−
−
−
−
−
−
−
−
−
−
8g
8h
−
−
−
−
−
−
a In ppm (δTMS = 0 ppm) at 125.7 and 62.9 (3d, 6b, f, h, 7b, d, f, h, and 8b, d, f, h) or 20.1 MHz (6a, d, 7a, and 8a). Solvent: CDCl3 or DMSO-d6
(for 7a, b), respectively.
b Assignments were supported by DEPT (except for 3d, 6a, 7b, f, h, and 8a, b, d), for 4d, 5a, 6c, 7g, and 8c also by HMQC and HMBC measurements.
c Further lines (see also Table 2): phenyl (6a), C-1: 138.3, C-2,6: 118.9, C-3,5: 127.9, C-4: 123.4, ArCH2: 28.7 (6–8c), OCH2: 64.1 (5–7e).
d Interchangeable assignments.
e Tosyl group.
f Doublet due to F,C-couplings, 1J: 242.5 (6f) and 246.2 (7h and 8h), 2J: 22.4 (6f) and 22.7 (7h and 9h), 3J: 8.4 (7h and 8h), 4J: 3.2 (8h).
g Two overlapping lines.
2.6.3. 3β-Acetoxy-21-chloromethylpregn-5-ene-20β-
N-4-tolylurethane (6b)
2.6.6. 3β-Acetoxy-21-chloromethylpregn-5-ene-20β-
N-4-ethoxyphenylurethane (6e)
6b (960 mg; 88%), mp 162–164 ◦C, Rf = 0.6 (ss E);
6e (980 mg; 85%), mp 157–159 ◦C, Rf = 0.50 (ss E);
20
20
[␣]D + 8 (c 1 in chloroform) (found C, 70.95; H, 8.05.
[␣]D − 50 (c 1 in chloroform) (found C, 69.36; H, 8.28.
C32H44O4NCl requires C, 70.89; H, 8.18%);
C33H46O5NCl requires C, 69.27; H, 8.10%).
2.6.4. 3β-Acetoxy-21-chloromethylpregn-5-ene-20β-
N-4-ethylphenylurethane (6c)
2.6.7. 3β-Acetoxy-21-chloromethylpregn-5-ene-20β-
N-4-chlorophenylurethane (6f)
6c (910 mg; 81%) mp 129–131 ◦C, Rf = 0.55 (ss E);
6f (900 mg; 80%), mp 116–118 ◦C, Rf = 0.70 (ss E);
20
20
[␣]D + 4 (c 1 in chloroform) (found C, 71.35; H, 8.22.
[␣]D + 11 (c 1 in chloroform) (found C, 66.27; H, 7.35.
C33H46O4NCl requires C, 71.26; H, 8.34%).
C33H41O4NCl2 requires C, 66.18; H, 7.35%).
2.6.5. 3β-Acetoxy-21-chloromethylpregn-5-ene-20β-
N-4-methoxyphenylurethane (6d)
2.6.8. 3β-Acetoxy-21-chloromethylpregn-5-ene-20β-
N-4-bromophenylurethane (6g)
6d (895 mg; 80%), mp 105–107 ◦C, Rf = 0.40 (ss E);
6g (950 mg; 78%), mp 140–142 ◦C, Rf = 0.60 (ss E);
20
20
[␣]D + 4 (c 1 in chloroform) (found C, 68.75; H, 8.02.
[␣]D + 13 (c 1 in chloroform) (found C, 61.52; H, 6.72.
C32H44O5NCl requires C, 68.86; H, 7.95%).
C31H41O4BrClN requires C, 61.34; H, 6.81%).

456
J. Wölfling et al. / Steroids 69 (2004) 451–460
2.6.9. 3β-Acetoxy-21-chloromethylpregn-5-ene-20β-
N-4-fluorophenylurethane (6h)
2.7.8. 17β-[3-(N-4-Chlorophenyl)tetrahydrooxazin-2-
on-6-yl]androst-5-en-3β-ol (7g)
6h (920 mg; 84%), mp 107–110 ◦C, Rf = 0.65 (ss E);
7g (915 mg; 94%), mp 179–183 ◦C, Rf = 0.30 (ss B);
20
20
[␣]D − 5 (c 1 in chloroform) (found C, 68.23; H, 7.67.
[␣]D − 70 (c 1 in chloroform) (found C, 72.06; H, 7.83.
C31H41O4NFCl requires C, 68.18; H, 7.57%).
C29H38O3NCl requires C, 71.96; H, 7.91%).
2.7.9. 17β-[3-(N-4-Fluorophenyl)tetrahydrooxazin-2-
on-6-yl]androst-5-en-3β-ol (7h)
2.7. 17β-[3-(N-Aryl)-tetrahydrooxazin-2-on-6-yl]androst-
5-en-3β-ol (7a–h)
7h (876 mg; 93%), mp 235–237 ◦C, Rf = 0.35 (ss B);
20
[␣]D − 69 (c 1 in chloroform) (found C, 74.63; H, 8.02.
2.7.1. General procedure
C29H38O3NF requires C, 74.49; H, 8.19%).
The individual compounds 6a–h (2 mmol) were dissolved
in methanol (30 ml), NaOCH3 (432 mg, 8 mmol) was added
and the solution was kept at the boiling point. The progress
of the reaction was monitored by TLC. Refluxing was con-
tinued until the starting material had been fully consumed
(about 120 min). The reaction mixture was then neutralized
with dilute HCl, and diluted with water. The resulting pre-
cipitate was filtered off, dissolved in dichloromethane and
chromatographed on silica gel with a mixture of ethyl ac-
etate/chloroform (5:95).
2.8. 17β-[3-(N-Aryl)tetrahydrooxazin-2-on-6-yl]androst-
4-en-3-one (8a–h)
2.8.1. General procedure
The individual compounds 7a–h (2 mmol) were dis-
solved in toluene (60 ml), aluminium isopropoxide (2.04 g,
10 mmol) and cyclohexanone (20 ml) were added, and the
mixture was stirred at the boiling point. The disappearance
of the starting material was followed by TLC. When the re-
action was complete, dilute H2SO4 (1%, 200 ml) was added
to the reaction mixture. Most of the organic solvent was
removed in vacuo and the residual emulsion was extracted
with chloroform. The organic phase was evaporated down
and the resulting crude product was chromatographed on
silica gel with ethyl acetate/chloroform (5:95).
2.7.2. 17β-[3-(N-Phenyl-tetrahydrooxazin-2-on-6-yl)]
androst-5-en-3β-ol (7a)
7a (786 mg; 86%), mp 212–214 ◦C, Rf = 0.45 (ss B);
20
[␣]D − 79 (c 1 in chloroform) (found C, 77.58; H, 8.57.
C29H39O3N requires C, 77.47; H, 8.74%).
2.7.3. 17β-[3-(N-4-Tolyl)tetrahydrooxazin-2-on-6-yl]
androst-5-en-3β-ol (7b)
2.8.2. 17β-[3-(N-Phenyl)tetrahydrooxazin-2-on-6-yl]
androst-4-en-3-one (8a)
7b (810 mg; 87%), mp 240–243 ◦C, Rf = 0.45 (ss B);
8a (720 mg; 80%), mp 234–235 ◦C, Rf = 0.55 (ss B);
20
[␣]D − 82 (c 1 in chloroform) (found C, 77.93; H, 9.05.
20
[␣]D + 46 (c 1 in chloroform) (found C, 77.64; H, 8.45.
C30H41O3N requires C, 77.71; H, 8.91%).
C29H37O3N requires C, 77.82; H, 8.33%).
2.7.4. 17β-[3-(N-4-Ethylphenyl)tetrahydrooxazin-2-on-
6-yl]androst-5-en-3β-ol (7c)
2.8.3. 17β-[3-(N-4-tolyl)tetrahydrooxazin-2-on-6-yl]
androst-4-en-3-one (8b)
7c (805 mg; 84%), mp 190–192 ◦C, Rf = 0.40 (ss B);
8b (685 mg; 74%), mp 208–209 ◦C, Rf = 0.45 (ss B);
20
[␣]D − 73 (c 1 in chloroform) (found C, 77.82; H, 8.98.
20
[␣]D + 45 (c 1 in chloroform) (found C, 77.94; H, 8.67.
C31H43O3N requires C, 77.95; H, 9.07%).
C30H39O3N requires C, 78.05; H, 8.52%).
2.7.5. 17β-[3-(N-4-Methoxyphenyl)tetrahydrooxazin-2-
on-6-yl]androst-5-en-3β-ol (7d)
2.8.4. 17β-[3-(N-4-Ethylphenyl)tetrahydrooxazin-2-
on-6-yl]androst-4-en-3-one (8c)
7d (790 mg; 82%), mp 232–234 ◦C, Rf = 0.35 (ss B);
8c (692 mg; 75%), mp 183–186 ◦C, Rf = 0.50 (ss B);
20
[␣]D − 81 ( c 1 in chloroform) (found C, 75.25; H, 8.55.
20
[␣]D + 52 (c 1 in chloroform) (found C, 78.37; H, 8.74.
C30H41O4N requires C, 75.12; H, 8.62%).
C31H41O3N requires C, 78.28; H, 8.69%);
2.7.6. 17β-[3-(N-4-Ethoxyphenyl)tetrahydrooxazin-2-on-
6-yl]androst-5-en-3β-ol (7e)
2.8.5. 17β-[3-(N-4-Methoxyphenyl)tetrahydrooxazin-2-
on-6-yl]androst-4-en-3-one (8d)
7e (832 mg; 84%), mp 207–210 ◦C, Rf = 0.45 (ss B);
8d (756 mg; 79%), mp 228–231 ◦C, Rf = 0.50 (ss B);
20
20
[␣]D + 2 (c 1 in chloroform) (found C, 75.55; H, 8.87.
[␣]D + 29 (c 1 in chloroform) (found C, 75.62; H, 8.25.
C31H43O4N requires C, 75.42; H, 8.78%).
C30H39O4N requires C, 75.44; H, 8.23%).
2.7.7. 17β-[3-(N-4-Bromophenyl)tetrahydrooxazin-2-on-
6-yl]androst-5-en-3β-ol (7f)
2.8.6. 17β-[3-(N-4-Ethoxyphenyl)tetrahydrooxazin-2-on-
6-yl]androst-4-en-3-one (8e)
7f (910 mg; 86%), mp 213–214 ◦C, Rf = 0.45 (ss B);
8e (872 mg; 88%), mp 198–201 ◦C, Rf = 0.45 (ss B);
20
20
[␣]D − 72 (c 1 in chloroform) (found C, 66.08; H, 7.32.
[␣]D + 126 (c 1 in chloroform) (found C, 75.61; H, 8.65.
C29H38O3NBr requires C, 65.90; H, 7.25%).
C31H41O4N requires C, 75.73; H, 8.41%).

J. Wölfling et al. / Steroids 69 (2004) 451–460
457
2.8.7. 17β-[3-(N-4-Bromophenyl)tetrahydrooxazin-2-
on-6-yl]androst-4-en-3-one (8f)
substituents. The cyclization typically occurred with neigh-
boring group participation, to yield N-phenyloxazolidone-2
or N-phenyltetrahydrooxazinone-2 derivatives, depending
on the starting compounds [14]. In the solvolysis of vici-
nal halourethanes, oxazolidinone derivatives condensed to
the estrane skeleton are formed, which are likewise poten-
tial pharmacons [15]. The cyclization of the four possible
isomeric 16-p-toluene-sulfonyloxymethyl-17-unsubstituted
phenylurethanes on the androstane and estrane skeletons
has already been investigated [16,17].
To prepare the ␣,␥-diol system on the side chain of the
sterane skeleton, we chose 3␤-acetoxypregn-5-en-20-one (1)
as starting compound. Elongation of the side-chain with a
hydroxymethyl group on C-21 in 1 was accomplished by
condensation with ethyl formate in the presence of NaOMe,
followed by reduction with KBH4 to the triols [10].
8f (874 mg; 83%), mp 222–225 ◦C, Rf = 0.60 (ss B);
20
[␣]D + 47 (c 1 in chloroform) (found C, 65.98; H, 6.67.
C29H36O3NBr requires C, 66.16; H, 6.89%).
2.8.8. 17β-[3-(N-4-Chlorophenyl)tetrahydrooxazin-2-
on-6-yl]androst-4-en-3-one (8g)
8g (790 mg; 81%), mp 210–213 ◦C, Rf = 0.40 (ss B);
20
[␣]D + 46 (c 1 in chloroform) (found C, 72.43; H, 7.41.
C29H36O3NCl requires C, 72.26; H, 7.53%).
2.8.9. 17β-[3-(N-4-Fluorophenyl)tetrahydrooxazin-2-
on-6-yl]androst-4-en-3-one (8h)
8h (780 mg; 83%), mp 237–239 ◦C, Rf = 0.55 (ss B);
20
[␣]D + 50 (c 1 in chloroform) (found C, 74.65; H, 7.91.
C29H36O3NF requires C, 74.81; H, 7.79%).
The simultaneous development of the new chiral centers
at C-20 permits the formation of 21-hydroxymethylpregn-
5-ene-3␤,20␤-diol (3a) and its 20␣ epimer (4a) in a ratio of
95:5. The separation of the two isomeric compounds in this
form was not possible. The cyclic acetonides of the mixture
of 21-hydroxymethyl-20-hydroxy compounds were acety-
lated in the 3␤ position, and the subsequent acidic hydrolysis
of the cyclic ketal produced a chromatographically separable
mixture of 3␤-acetoxy-21-hydroxymethylpregn-5-en-20-ol
isomers (3d and 4d). The separation of the two isomeric
compounds in this form was possible. All 13C NMR chemi-
cal shifts of the epimeric pair 3d and 4d are the same within
0.5 ppm. Only the shifts of C-14, C-17, and C-18, carbons
near to the chiral center (in position C-20) with different
configurations, differ by 0.7, 0.6, and 0.8 ppm, respectively.
Similarly, the 1H NMR chemical shifts of the diastereomers
are practically identical (ꢀδ ≤ 0.02 ppm). The only excep-
tion is the 18-CH3 group, which is similar close to C-20,
the shifts of which differ by 0.09 ppm for the isomers.
For the further reactions, we used only the pure 20␤ iso-
mer (3d). In the earlier literature it was supposed that the
reduction of the C-20 ketone is stereoselective and that C-20
has the R configuration [10].
The selective p-tosyl ester formation from 3d re-
sulted in 5a, which, in a subsequent reaction with phenyl
isocyanate, was converted to 3␤-acetoxy-21-p-toluene-
sulfonyloxymethylpregn-5-ene-20␤-phenylurethane (6a).
For the preparation of substituted phenylurethane deriva-
tives 6b–h, the starting compound was 3␤-acetoxy-21-
chloromethylpregn-5-en-20␤-ol (5b), which was prepared
by the Appel reaction of 3d [18]. The OH → Cl change
in 5b as compared with 3d is confirmed by the very large
upfield shift (by 19.1 ppm) of the 13C NMR line of the
neighboring carbon. 5b Reacted with phosgene in the pres-
ence of triethylamine to furnish the desired 3␤-acetoxy-
21-chloromethylpregn-5-ene-20␤-chlorocarbonic acid ester.
Without isolation, this compound was reacted with the
appropriately substituted anilines to yield 6b–h. The pres-
ence of the O-tosyl group in 5a and 6a is demonstrated
by the characteristic spectral data on this substituent (e.g.
2.9. Determination of 5α-reductase activity and its
inhibition in the human prostate
Human prostates were obtained surgically from patients
with benign prostatic hyperplasia and homogenized [11].
The protein content was determined according to Lowry
et al. [12]. The homogenate was incubated with 200 nM
[4-14C]testosterone and the test compound (in four differ-
ent concentrations) in the presence of 1 mM NADPH for
20 min at 37 ◦C. The control experiments were performed
without the test substances in every series. The enzymatic
reaction was stopped by the addition of ethyl acetate and
freezing. After extraction the radioactive testosterone and
dihydrotestosterone were separated by TLC and a Packard
Radiochromatogram Scanner was used to trace the separated
steroids. Following measurement of the radioactivity of the
dihydrotestosterone formed and the substrate testosterone,
the 5␣-reductase activity was calculated and expressed in
terms of pmol dihydrotestosterone per mg of protein for a
20-min incubation. The inhibitory effects of the compounds
investigated are given in terms of IC50 values, i.e. the con-
centration of inhibitor at which the 5␣-reductase activity was
decreased to 50%.
3. Results and discussion
3.1. Synthetic studies
For the synthesis of tetrahydrooxazinone-2 ring at po-
sition 17␤ of the sterane ring, we chose the basic cy-
clization of the ␣,␥-haloalkyl-N-arylurethanes in alkaline
media. Scott et al. examined the ring-closure reactions of
␣,␤- and ␣,␥-haloalkyl N-arylurethanes in alkaline me-
dia [13]. They found that the N-arylurethane group and
its substituted derivatives reacted at different rates, de-
pending on the inductive and conjugative effects of the

458
J. Wölfling et al. / Steroids 69 (2004) 451–460
the ␯SO2 IR bands, the methyl singlet with 3H intensity
and the AAꢀBBꢀ multiplet of the p-disubstituted aromatic
of the deprotonated acid amide. This cyclization process is
typical neighboring group participation. In the notation pro-
posed by Winstein and Boschan [14], the process can be
characterized by the symbol (N− − 6).
1
ring in the H NMR spectra and the 13C NMR lines of the
methyl and aromatic carbons in the 13C NMR spectra) and
also by the downfield shifts (by 0.29 and 0.4 ppm) of the
The 3-acetoxy group in 3d, 4d, 5a, b, and 6a–h gives a
1
high ␯C O frequency (1713–1740 cm−1) in the IR spectrum,
=
21-methylene signal in the H NMR spectra. The urethane
carbonyl line appears in the expected interval [19], between
152.3 and 154.1 ppm, for all compounds in the series 6–8.
The urethane NH in series 5 gives a broad signal of 1H
intensity in all (a–h) cases (6.6–7.0 ppm).
Compounds 6a–h were subjected to methanolysis in the
presence of four equivalents of NaOCH3. Under these exper-
imental conditions, N-phenyl-tetrahydrooxazinone-2 rings
(7a–h) were formed in rapid reactions. The cyclization can
be explained by the nucleophilic attack of the nitrogen atom
expected for saturated esters [20], a 1H NMR methyl signal
(2.03 0.01 ppm) and a 13C NMR carbonyl line (170.8
0.4 ppm).
Oppenauer oxidation of 7a–h resulted in good yields of
the 17␤-[3-(N-aryl)tetrahydrooxazin-2-on-6-yl]androst-4-
en-3-ones 8a–h.
The change OAc → OH causes a moderate upfield shift
of the C-3 line, which appears at 74.2 0.2 ppm (series 6
and 3d, 4d, and 5, a, b) or at 71.9 0.3 ppm (series 7). This
CH3
HC OH
C
O
HC
C O
AcO
1
2
H2C OR2
CH2
H2C OH
CH2
R1 R2 R3
3
R3O
C H
H C OH
H
H
H
a
b
c
d
C(CH3)2
H
Ac
Ac
C(CH3)2
H
H
R1O
R1O
4d (20α)
3 (20β)
R
H2C R1
H2C
CH OX
N
O
O
R
6-8
phenyl
a
b
c
d
e
f
p-tolyl
p-C2H5-phenyl
p-CH3O-phenyl
AcO
HO
7
5 (X = H)
6 (X = CONHR)
p-C2H5O-phenyl
p-Br-phenyl
p-Cl-phenyl
p-F-phenyl
R
N
O
g
h
R
R
1 = OTs (5a, 6a)
1 = Cl (5b, 6b - h)
O
O
8
Scheme 1.

J. Wölfling et al. / Steroids 69 (2004) 451–460
459
Table 4
unsubsituted N-phenyl compound 8a is 0.20. As concerns
the effects of substituents at position 4 of the phenyl ring in
8a, the introduction of an ethyl (8c) or ethoxy (8e) group
resulted in a weak enhancement of 5␣-reductase inhibition.
Substitution with halogens (8f and 8g), methyl (8b) or
methoxy (8d) caused a drop-off in inhibition ability.
In vitro inhibition of 5␣-reductase by steroidal tetrahydrooxazin-2-ones
(8a–g)
Compounds
IC50 value (nM)
Relative inhibition
IC50finasteride IC50 compound
Finasteride
55
270
>1000
260
1
8a
8b
8c
8d
8e
8f
0.20
<0.055
0.21
0.09
0.22
0.11
0.13
Acknowledgements
600
245
500
420
The authors thank the Hungarian Scientific Research Fund
(OTKA T-042673, T-043634, and TS-040732) for financial
support.
8g
line is dramatically shifted downfield, of course, for series 8,
where it appears at 199.5 0.5 ppm, values characteristic of
ketones [21a]. The H-3 signal is also shifted downfield (due
to the −IS effect of the O-acetyl group) by ca. 1.2 ppm, as
expected [21b] for series 5 as compared with series 7 and 8.
The polarization of the enone moiety leads [21c] to a very
great difference (>47 ppm) between the C-4 and C-5 shifts
(δC-4: 124.0 0.3 and δC-5: 171.3 0.4) in series 8.
References
[1] Tapolcsányi P, Wölfling J, Schneider Gy. The preparation of the
four stereoisomers of 16-hydroxymethyl-5␣-androstane-3␤,17-diol.
Steroids 2001;66:623–35. Part 14.
[2] Hall H. Cytochrome P-450 C21scc: one enzyme with two actions:
hydroxylase and lyase. J Steroid Biochem Mol Biol 1991;40:527–32.
[3] Njar VCO, Kato K, Nnane IP, Grigoryeev DN, Long BJ,
Brodie AMH. Novel 17-Azo-yl steroids, potent inhibitors
of human cytochrome 17␣-hydroxylase-C17,20-lyase (P45017␣):
potential agents for the treatment of prostate cancer. J Med Chem
1998;41:902–12.
[4] Jarman M, Barrie SE, Llera JM. The 16,17-double bond is
needed for irreversible inhibition of human cytochrome P45017␣
by abiraterone (17-(3-pyridyl)androsta-5,16-dien-3␤-ol) and related
steroidal inhibitors. J Med Chem 1998;41:5375–81.
[5] Guarna A, Occiato EG, Mechetti F, Marrucci A, Danza G, Serio M,
Paoli P. 19-Nor-10-azasteroids, a new class of steroid 5␣-reductase
inhibitors: 2. X-ray structure, molecular modeling, conformational
analysis of 19-nor-10-azasteroids and comparison with 4-azasteroids
and 6-azasteroids. J Med Chem 1997;40:3466–77.
1
The H and 13C NMR signals of the19-CH3 group are
sensitive to the change of the ꢀ5-double bond (series 6 and
7) to the ꢀ4-en-3-one moiety in series 8: δH(19-CH3) is
1.00 0.03 for series 6 and 7, but 1.19–1.20 for series 8,
while δC(19-CH3) is 20.0 0.8 for series 6 and 7, but 17.6
0.2 for series 8. The freely rotating aryl-urethane group
in series 6 can lead to a shielding effect of ca. 0.1 ppm on
the 18-CH3 hydrogen relative to series 7 and 8 (where the
ring structure forces the aryl group far from 18-CH3) due to
the anisotropic neighboring effect of the benzene ring [21d].
The spectral data characteristic of the substituents (Me, Et,
OMe, OEt, halogens) on the aryl group are observed in series
6–8 ( Tables 1–3; Scheme 1).
[6] Li J, Li Y, Son C, Banks P, Brodie A. 4-Pregnen-3-one-20␤-
carboxaldehyde: a potent inhibitor of 17␣-hydroxylase/C17,20-lyase
It is interesting to note that in the more rigid ring B of
the steroidal skeleton in series 6 and 7 (caused by the ꢀ5-
double bond) the C-9 line is shifted upfield by ca. 3.5 ppm as
compared with series 8, probably because there is no steric
hindrance in the latter series (field effect) [21e,22].
and of 5␣-reductase.
313–20.
J Steroid Biochem Mol Biol 1992;42:
[7] Guarna A, Occhiato EG, Machetti F, Giacomelli V. 19-Nor-10-
azasteroids. 5. A synthetic strategy for the preparation of (+)-17-(3-
pyridyl)-(5␤)-10-azaestra-1,16 dien-3-one, a novel potential inhibitor
for human cytochrome P45017␣ (17␣-hydroxylase/C17,20-lyase). J
Org Chem 1999;64:4985–9.
[8] Russel DW, Wilson JD. Steroid 5␣-reductase: two genes/two
enzymes. Ann Rev Biochem 1994;6325–61.
[9] Normington K, Russel DW. Tissue distribution and kinetic
characterisation of rat steroid 5␣-reductase isoenzymes. J Biol Chem
1992;267:19548–54.
[10] Hirsch AF, Fujimoto GI. Synthesis of 21-hydroxymethylprogesterone.
J Org Chem 1970;35:495–8.
[11] Faredin I, Toth I, Wölfling J, Schneider Gy, Mesko E. In vitro
inhibitory effects o 16-methyl-substituted steroids on 5␣-reductase
in rat and human prostates. Steroids 1994;59:568–71.
3.2. Effects of steroidal tetrahydrooxazin-2-ones on
5α-reductase activity
The inhibitory effects of compounds 8a–g on human
prostatic 5␣-reductase were investigated with an in vitro
incubation technique. Finasteride (17␤-(N-tert-butylcarba-
moyl)-4-aza-5␣-androst-1-en-3-one), a potent 5␣-reductase
inhibitor already applied in medical practice, was used as a
reference compound.
[12] Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ. Protein
measurements with the Folin phenol reagent.
1951;193:265–75.
J Biol Chem
IC50 values and relative inhibitory efficiencies as com-
pared with the IC50 of finasteride are presented in the
Table 4. The investigated steroids (8a–g) exhibit some 5␣-
reductase inhibitory effect. The IC50 values of 8a–g vary
between 270 and 600 nM, with the exception of 8b, which
is a very weak inhibitor. The relative inhibitory effect of the
[13] Scott FL, Glick RE, Winstein S. Amido, ureido and urethano
neighbouring group participation. Experientia 1957;13:183–5.
[14] Winstein S, Boschan R. The role of neighboring groups in
replacement reactions. XVII. Complex neighboring groups. The
benzamido group. J Am Chem Soc 1950;72:4669–77.

460
J. Wölfling et al. / Steroids 69 (2004) 451–460
[15] Ponsold K, Grosse P. Ringschlußreaktionen vicinaler halogen–
urethane der steroidreihe. J f prakt Chemie 1981;323:801–18.
[16] Schneider G, Hackler L, Sohár P. Steroids, XXIX. Neighbouring
group participation, VI. N-Arylurethane neighbouring group
participation in the 16-hydroxymethylandrost-5-ene-3,17-diol series.
Acta Chim Acad Sci Hung 1982;111:285–96.
[17] Hackler L, Schneider G, Sohár P. N-Arylurethane neighbouring
group participation during solvolysis of 3-methoxy-17-N-
phenylcarbamoyloxy-16-p-tolylsulfonyloxymethyl-estra-1,3,5(10)-
triene stereoisomers. Tetrahedron 1992;42:9305–14.
[18] Appel R, Halstenberg M. In: Cadogan JIG, editor. Organophosphorus
reagents in organic synthesis. Academic Press: New York;
p. 378–424.
[19] Levy GC, Nelson GL. Carbon-13 nuclear magnetic resonance for
organic chemists. New York: Wiley; 1972. p. 126.
[20] Holly S, Sohár P. Theoretical and technical introduction to the series
(1979).In: Láng L, Prichard, WH, editors. Absorption spectra in the
infrared region. Budapest: Akadémiai Kiadó; 1975. p. 94.
[21] Sohár P. Nuclear magnetic resonance spectroscopy, vol. 2. Boca
Raton, Florida: CRC Press; 1983. p. 180–185Sohár P. Nuclear
magnetic resonance spectroscopy, vol. 2. Boca Raton, Florida: CRC
Press; 1983. p. 28Sohár P. Nuclear magnetic resonance spectroscopy,
vol. 2. Boca Raton, Florida: CRC Press; 1983. p. 181Sohár P. Nuclear
magnetic resonance spectroscopy, vol. 1. Boca Raton, Florida:
CRC Press; 1983. p. 35–38Sohár P. Nuclear magnetic resonance
spectroscopy, vol. 2. Boca Raton, Florida: CRC Press;1983. p.
154–155
[22] Grant DM, Cheney BV. Carbon-13 magnetic resonance. VII. Steric
perturbation of the carbon-13 chemical shift. J Am Chem Soc
1967;89:5315–8.