Microbial transformation of 3-hydroxy-5,6-cyclopropanocholestanes - An alternative route to 6-methylsteroids

Article abstract of DOI:10.1016/S0039-128X(00)00170-7

Incubation of 3β-hydroxy-5,6α-cyclopropano-5α-cholestane (4), 3β-hydroxy-5,6β-cyclopropano-5β-cholestane (5), and 3β-hydroxy-5,6α-cyclopropano-5α-cholest-7-ene (6) with Mycobacterium sp. (NRRL B-3805) gave a mixture of side chain cleaved 17-keto steroids as the major products in 52, 57, and 69% yields, respectively. Among these 17-keto steroids, the cyclopropyl ring eliminated product, androst-4-ene-3,17-dione (9), was isolated in 6, 4, and 8% yields, respectively. A cyclopropyl ring migration product, 6α,7α-cyclopropanoandrost-4-ene-3,17-dione (16), was isolated from the incubation mixture of 6 in 4% yield, also 10% yield of 16 was obtained when 5,6α-cyclopropano-5α-androst-7-ene-3,17-dione (12) was incubated. The cyclopropyl ring opening and subsequent reduction followed by oxidation of the two major biotransformation products, 5,6β-cyclopropano-5β-androsta-3,17-dione (10) and 5,6α-cyclopropano-5α-androsta-3,17-dione (7), gave 6β- and 6α-methylandrost-4-ene-3,17-dione in 60, and 45% yields, respectively. (C) 2000 Elsevier Science Inc.

Full text of DOI:10.1016/S0039-128X(00)00170-7

Steroids 65 (2000) 863–870
Microbial transformation of 3-hydroxy-5,6-cyclopropanocholestanes -
An alternative route to 6-methylsteroids
Jiann-Long Yan, Shoei-Sheng Lee, K.C. Wang*
School of Pharmacy, College of Medicine, National Taiwan University, No. 1, Section 1, Jen-Ai Road, Taipei 100, Taiwan, Republic of China
Received 15 May 2000; accepted 16 June 2000
Abstract
Incubation of 3␤-hydroxy-5,6␣-cyclopropano-5␣-cholestane (4), 3␤-hydroxy-5,6␤-cyclopropano-5␤-cholestane (5), and 3␤-hydroxy-
5,6␣-cyclopropano-5␣-cholest-7-ene (6) with Mycobacterium sp. (NRRL B-3805) gave a mixture of side chain cleaved 17-keto steroids as
the major products in 52, 57, and 69% yields, respectively. Among these 17-keto steroids, the cyclopropyl ring eliminated product,
androst-4-ene-3,17-dione (9), was isolated in 6, 4, and 8% yields, respectively. A cyclopropyl ring migration product, 6␣,7␣-cyclopro-
panoandrost-4-ene-3,17-dione (16), was isolated from the incubation mixture of 6 in 4% yield, also 10% yield of 16 was obtained when
5,6␣-cyclopropano-5␣-androst-7-ene-3,17-dione (12) was incubated. The cyclopropyl ring opening and subsequent reduction followed by
oxidation of the two major biotransformation products, 5,6␤-cyclopropano-5␤-androsta-3,17-dione (10) and 5,6␣-cyclopropano-5␣-andro-
sta-3,17-dione (7), gave 6␤- and 6␣-methylandrost-4-ene-3,17-dione in 60, and 45% yields, respectively. © 2000 Elsevier Science Inc. All
rights reserved.
Keywords: Microbial transformation; 5,6-Cyclopropanocholestanes; Mycobacterium sp. (NRRL B-3805); 5,6-Cyclopropanoandrosta-3,17-diones;
6-Methylandrost-4-ene-3,17-diones
1. Introduction
dione (3) as the major product in 34 and 35% yields,
respectively [6,7] (Scheme 1).
It is well documented that a wide range of microorgan-
isms are capable of utilizing sterols such as cholesterol or
phytosterols as the sole carbon source for growth and prop-
agation [1]. It is also well known that in the sterol degrading
microorganisms, the ring nucleus and side chain are metab-
olized by different enzyme systems simultaneously and in-
dependently [2]. The microorganism, Mycobacterium sp.
(NRRL B-3805) is a mutant developed by Marsheck and
coworkers for selective cleavage of the sterol side chain to
give C19 and other useful steroid intermediates [3]. The
brilliant work of Sih and coworkers clarified the pathways
and mechanisms involved in the microbial degradation of
steroid side chains [4,5].
It is interesting to find out whether this cyclopropyl ring
opening capability of M. sp. (NRRL B-3805) would be
applicable to 5,6-cyclopropanosterols. Therefore, 3␤-hy-
droxy-5,6␣-cyclopropano-5␣-cholestane (4) [8,9], 3␤-hy-
droxy-5,6␤-cyclopropano-5␤-cholestane (5) [8,10], and 3␤-
hydroxy-5,6␣-cyclopropano-5␣-cholest-7-ene (6) were
prepared and incubated with this microorganism. In this
report, we like to describe the results obtained in this bio-
transformations and also utilization of the major products
isolated for the preparation of 6-methylandrost-4-ene-3,17-
diones.
In previous reports, we disclosed that the cyclopropyl
ring in cycloartenol (1) and 24-methylenecycloartanol (2)
were opened to form the C-19 methyl by M. sp. (NRRL
B-3805) with the isolation of androsta-4,8(14)-diene-3,17-
2. Experimental
2.1. General methods
All of the solvents and inorganic chemicals were reagent
grade. Cholesterol and 7-dehydrocholestesterol were pur-
chased from Sigma Chem. Co., St. Louis, Mo. Aluminum
isopropoxide, diethylzinc, and diiodomethane were pur-
* Corresponding author. Tel.: ϩ886-2-2312-3456 ext. 8383, ϩ886-2-
2393-0576; fax: ϩ886-2-2391-1307.
0039-128X/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved.
PII: S0039-128X(00)00000-0

864
J.-L. Yan / Steroids 65 (2000) 863–870
benzoic acid (5.1 g, 20.7 mmol) for 3 h with stirring.
Termination of reaction with 20% NaHSO3 solution (13 g,
25 mmol), addition of saturated NaHCO3 solution (50 ml),
extraction with CH2Cl2 (50 ml x 3), and removal of the
CH2Cl2 gave crude mixture (12.3 g). Separation of this
crude mixture (12.3 g) by Si gel column (200 g) eluted with
n-C6H12:EtOAc ϭ 5:1 mixture gave crude cholesterol 4,5-
epoxide (6.2 g), 4 (0.95 g), and 5 (2.6 g). Further purifica-
tion with small Si gel column and recrystallizations gave
pure 4 (0.61 g, 5.8%) and 5 (1.83 g, 17.7%), respectively.
Compound 4 and 5 obtained were identical in every respects
(mp, IR and 1H-NMR) to 4 [8,9] and 5 [8,10] reported in the
literatures. The structure of 4 was further confirmed by
X-ray diffraction analysis.
Scheme 1.
chased from Aldrich Chem. Co., Inc., Milwaukee, WI.
Agar, dextrose, nutrient broth, and yeast extract were from
Difco Lab., Detroit, MI. Mercuric trifluoroacetate was ob-
tained from Acros Organics, Geel, Belgium. Lithium alu-
minum hydride, Normal phase silica gels of 7734 (70–230
mesh) and 9385 (230–400 mesh) for column chromatogra-
phy, and TLC plates of 573 (Si60 with F254, 0.25 mm) were
purchased from E. Merck A. G., Darmstadt, Germany.
Melting points measured on a Buchi 510 melting point
apparatus were uncorrected. Optical rotations were mea-
sured on a Jasco DIP-370 digital polarimeter. UV spectra
were recorded on a Hitachi 150–20 double beam spectrom-
eter. IR spectra were recorded on a Jasco A-100 infrared
spectrophotometer. 1H- and 13C-NMR spectra were ob-
tained on Bruker AM-300 or Bruker AMX-400 spectropho-
tometer in CDCl3 or CD3OD using the residual solvent
peaks as reference standards. EIMS were recorded on Finni-
gan Mat 4500 GC/MS and HREIMS on JEOL JMX-HX-
100 mass spectrometer.
2.3.2. 3␤-Hydroxy-5,6␣-cyclopropano-5␣-cholest-7-ene
(6) [11]
A dry solution of 7-dehydrocholesterol (2.0 g, 5.2 mmol)
in 1,2-dichloroethane (80 ml) was cooled to 0°C. To this
solution, CH2I2 (0.9 ml, 11 mmol) then Et2Zn (20.7 ml of
1.1M toluene solution, 22 mmol) were added successively
dropwise with stirring and the reaction mixture was kept at
10°C for 120 h. At the end of reaction period, saturated
NH4Cl solution (50 ml) and H2O (50 ml) were added and
extracted with CHCl3 (100 ml x 3). Work-up of the com-
bined CHCl3 extract gave crystalline residue (1.69 g). The
residue was purified by a Si gel column (60.0 g, 230–400
mesh) eluted with CHCl3 and recrystallization to give pure
6 (1.5 g, 72.4%).
2.2. Maintenance and growth of microorganism
6: mp 158–159°C; [␣]D25-140 (c 0.5, CHCl3); IR (KBr)
␯ max 3475 (OH), 2925, 2860, 1470, 1450, 1375, 1360,
1320, 1280, 1220, 1165, 1130, 1070, and 1050 cmϪ1; 1H-
NMR ␦ (CDCl3) 5.44 (1 H, ddd, J ϭ 2.2, 2.5, 4.9 Hz, H-7),
3.75 (1 H, tt, J ϭ 4.8, 11.2 Hz, H-3␣), 2.12 (1 H, ddd, J ϭ
2.2, 11.2, 12.3 Hz, H-4␤), 2.01 (1 H, ddd, J ϭ 3.0, 4.2, 12.6
Hz, H-16␤), 1.06 (1 H, ddd, J ϭ 3.4, 4.9, 7.7 Hz, H-6␤),
1.02 (3 H, s, H-19), 0.88 (3 H, d, J ϭ 6.6 Hz, H-21), 0.844
(3 H, d, J ϭ 6.6 Hz) and 0.839 (3 H, d, J ϭ 6.6 Hz) (H-26
and H-27), 0.73 (1 H, br dd, J ϭ 3.4, 3.7 Hz) and 0.42 (1 H,
dd, J ϭ 3.7, 7.7 Hz) (H-28’s), 0.51 (3 H, s, H-18); 13C-NMR
␦ (CDCl3) [12] 135.5 (s, C-8), 121.1 (d, C-7), 70.1 (d, C-3),
55.9 (d, C-17), 54.6 (d, C-14), 42.7 (t, C-4), 42.2 (s, C-13),
41.8 (d, C-9), 39.5 (t, C-24), 39.2 (t, C-16), 36.2 (d, C-20),
36.1 (t, C-22), 34.9 (t, C-1), 32.3 (s, C-10), 32.1 (t, C-2),
29.1 (s, C-5), 28.0 (d, C-25), 27.9 (t, C-12), 23.9 (t, C-23),
23.5 (t, C-16), 22.5 (q) and 22.8 (q) (C-26 and C-27), 21.5
(t, C-11), 19.2 (t, C-28), 18.8 (q, C-21), 17.9 (d, C-6), 16.8
(q, C-19), 11.7 (q, C-18); NOE data: H-18 (␦ 0.51) to H-21
(2.3%), H-19 (7.0%) and H-16␤ (2.2%), H-19 (␦ 1.02) to
H-4␤ (11.2%) and H-18 (6.1%); HREIMS m/z [M]ϩ
398.3542 (calcd for C28H46O 398.3549); EIMS m/z [M]ϩ
398 (100), 383 (34), 365 (51), 328 (30), 313 (33), 299 (28),
285 (52), 267 (46), 259 (26), 245 (30), 225 (41), 213 (35),
199 (28), 173 (35), 163 (51), 144 (98), 133 (32), 115 (45);
Crystal data: Monoclinic P21; a ϭ 12.1189 (1) Å, alpha ϭ
Mycobacterium sp. (NRRL B-3805) was maintained on a
Bacteria agar (Dextrose 1%, Yeast extract 1%, Nutrient
broth 0.8%, and Agar 3.5%) at 26°C for 11 days then
transferred and grown in Nutrient broth-dextrose medium
(Nutrient broth 1.6% and Dextrose 4%) at 26–28°C on a
rotary shaker (250 rpm, 1-in stroke) for 48 h and used as
inoculum. Transformations were carried out in 2-L Erlen-
meyer flasks containing 400 ml of the same medium and
incubated for 6 days.
2.3. Preparation of substrates
2.3.1. 3␤-Hydroxy-5,6␣-cyclopropano-5␣-cholestane (4),
and 3␤-Hydroxy-5,6␤-cyclopropano-5␤-cholestane (5) [9]
To a solution of cholesterol (10 g, 25.9 mmol) in 1,2-
dichloroethane (100 ml) was added with stirring 15% Et2Zn
in toluene (94 ml, 103.5 mmol) and CH2I2 (16.7 ml, 207
mmol) under nitrogen. The mixture was stirred at room
temperature for 7 days, then saturated NH4Cl solution (100
ml) was added to it and filtered to remove yellow precipi-
tate. The filtrate was extracted with CH2Cl2 (100 ml x 3),
and the combined extract was dried over Na2SO4, evapo-
rated under reduced pressure to give yellow-colored solid
(10.7 g). This solid obtained was dissolved in CH2Cl2 (200
ml), cooled in an ice-bath, and reacted with m-chloroper-

J.-L. Yan / Steroids 65 (2000) 863–870
865
90°, b ϭ 6.1078 (1) Å, beta ϭ 95.558 (1)°, c ϭ 17.2301 (3)
Å, gamma ϭ 90°; volume ϭ 1269.37 (3) Å, Z ϭ 2. Intensity
data were collected on a CAD-4 diffractometer with ␪/2␪
scan mode, using monochromated MoK␣ radiation. Data
were measured up to 2 ␪ of 54.96°. A total of 11050
reflections were collected. Among them, 5816 were consid-
ered to be independent reflections ( Ͼ 2.05 ␴ (I)). Final
agreement indices are R(F) ϭ 0.0859, WR(F) ϭ 0.2340,
GoF ϭ 0.997, based on anisotropic refinement of all non-
hydrogen atoms.
NMR) to 5,6␤-cyclopropano-5␤-androsta-3,17-dione re-
ported in the literature [13].
25
11: mp 163–164°C; [␣]D ϩ21 (c 1.0, CHCl3); IR
(KBr) ␯ max 3500 (OH), 2950, 2850, 1725, 1455, 1379,
1260, 1210, 1130, 1110, 1100, 1050 and 1020 cmϪ1; 1H-
NMR ␦ (CDCl3) 4.03 (1 H, m, H-3), 2.41 (1 H, dd, J ϭ 8.9,
19.2 Hz) and 2.05 (1 H, dt, J ϭ 19.2, 9.0 Hz) (H-16’s), 1.94
(1 H, m) and 0.86 (1 H, dd, J ϭ 3.2, 14.8 Hz) (H-4’s), 1.75
(1 H, dt, J ϭ 12.6, 3.2 Hz, H-12␤), 1.06 (1 H, m, H-6), 0.84
(3 H, s, H-19) and 0.81 (3 H, s, H-18), 0.24 (1 H, dd, J ϭ
4.8, 5.2 Hz) and -0.03 (1 H, dd J ϭ 4.8, 8.8 Hz) (H-20’s);
13C-NMR ␦ (CDCl3) 221.1 (s, C-17), 67.8 (d, C-3), 52.0 (d,
C-14), 47.7 (d, C-9), 47.5 (s, C-13), 42.9 (t, C-4), 35.8 (t,
C-16), 34.3 (s, C-10), 33.4 (t, C-1), 31.6 (t, C-12), 30.1 (d,
C-8), 29.44 (t, C-7), 29.40 (t, C-2), 22.9 (s, C-5), 22.1 (q,
C-19), 21.8 (t, C-15), 21.3 (t, C-11), 21.1 (d, C-6), 13.6 (q,
C-18), 11.6 (t, C-20); HREIMS m/z [M]ϩ 302.2248 (calcd
for C20H30O2 302.2246); EIMS m/z [M]ϩ 302 (20), 284
(100), 269 (83), 256 (14), 243 (10), 227 (14), 199 (8), 187
(9), 173 (14), 159 (19), 150 (72), 133 (33), 119 (31), 105
(43); Crystal data: Orthorhombic P212121; a ϭ 20.4700 (3)
Å, alpha ϭ 90°, b ϭ 7.34320 (10) Å, beta ϭ 90°, c ϭ
11.2524 (2) Å, gamma ϭ 90°, volume ϭ 1691.41 (5) Å3,
Z ϭ 4. Intensity data were collected on a CAD-4 diffrac-
tometer with ␪/2␪ scan mode, using monochromated MoK␣
radiation. Data were measured up to 2 ␪ of 55.0°. A total of
20232 reflections were collected. Among them, 3884 were
considered to be independent reflections ( Ͼ 2.05 ␴ (I)).
Final agreement indices are R(F) ϭ 0.0528, WR(F) ϭ
0.1464, GoF ϭ 1.096, based on anisotropic refinement of all
non-hydrogen atoms.
2.4. Biotransformation of 4 and isolation of products
Compound 4 (0.36 g), mp 159–160°C (lit. 159–160°C)
[8,9] dissolved in DMF (7.5 ml) was distributed evenly
among 3 2-L Erlenmeyer flasks. At the end of incubation
period, the culture broth was acidified with HOAc to a pH
of 3.0 and extracted with CHCl3 (0.4 L x 3). The combined
CHCl3 layer was washed with distilled H2O and dried over
Na2SO4 then evaporated to dryness in vacuo to give a
brownish semisolid residue (0.68 g). The residue was chro-
matographed over a Si gel column (100 g) and eluted with
CHCl3, and then with CHCl3-Me2CO mixture. In the frac-
tion eluted with CHCl3, 7 (0.117 g, 43%) was obtained after
purification by rechromatographed over another Si gel col-
umn (20 g) and crystallization. In the fraction eluted with
CHCl3-Me2CO (99:1), 9 (0.06 g, 6%) was obtained after
purification by rechromatographed over another Si gel col-
umn (10 g) and crystallization. In the fractions eluted with
CHCl3-Me2CO (97:3), 8 (0.010 g, 3.6%) was obtained after
purification by rechromatographed over another Si gel col-
umn (8 g) and crystallization.
2.6. Incubation of 7 and 10
25
Compound 7 with mp 216–217°C; [␣]D ϩ36 (c 1.0,
CHCl3), and Compound 8 with mp. 178–179°C; [␣]D25 -13
(c 1.0, CHCl3) were identical in every respects (mp, IR, and
1H-NMR) to 5,6␣-cyclopropano-5␣-androsta-3,17-dione
(i) Compound 7 (100 mg) dissolved in DMF (3 ml) was
incubated in one 2-L Erlenmeyer flasks for 120 h. Work-up
of the incubation mixture led to the isolation of 9 (6.2 mg,
6%) and recovery of 7 (70.1 mg, 70%). (ii) Incubation of 10
(100 mg) as mentioned above led to the isolation of 9 (4.1
mg, 4%) and recovery of 10 (70.2 mg, 70%).
and
3␤-hydroxy-5,6␣-cyclopropano-5␣-androsta-17-one
reported in the literature [13], respectively. Compound 9
was found to be identical in every respect (mmp, IR, UV
and 1H-NMR) to an authentic sample of androst-4-ene-3,17-
dione (Aldrich, Milwaukee, W1, U.S.A.)
2.7. Biotransformation of 6 and isolation of products
2.5. Biotransformation of 5 and isolation of products
Compound 6 (1.2 g), dissolved in DMF (30 ml) was
incubated in 12 2-L Erlenmeyer flasks as mentioned above.
Work-up of the combined CHCl3 extract gave a brownish
semisolid residue (2.01 g). Chromatography of the residue
over a Si gel column (600 g) gave the following result:
compound 12 (90 mg, 10%) from CHCl3 eluted fractions;
Compound 9 (69 mg, 8%) from CHCl3-Me2CO (96:4)
eluted fractions; Compound 16 (36 mg, 4%) from CHCl3-
Me2CO (94:6) eluted fractions; Compound 13 (370 mg,
41%) from CHCl3-Me2CO (95:5) eluted fractions; Com-
pound 14 (38 mg, 4%) from CHCl3-Me2CO (93:7) eluted
fractions; and Compound 15 (20 mg, 2%) from CHCl3-
Me2CO (92:8) eluted fractions.
Compound 5 (1.2 g), mp 109–110°C (lit, 108–110°C)
[8,10], dissolved in DMF (30 ml) was incubated in 12 2-L
Erlenmeyer Flasks. Acidification, extraction, and work-up
of the incubation mixture gave a brownish semisolid residue
(1.98 g), The residue was chromatographed over a Si gel
column (300 g) and eluted with solvents as mentioned
above. Compound 10 (0.348 g, 38%), 9 (0.035 g, 4%), and
11 (0.136 g, 15%) were obtained respectively after purifi-
cation and crystallization.
Compound 10 with mp 106–107°C, [␣]D25 ϩ153 (c 1.0,
CHCl3) was identical in every respects (mp, IR, and H-
1

866
J.-L. Yan / Steroids 65 (2000) 863–870
12: mp 210–213°C; [␣]D25-63 (c 0.5, CHCl3); IR (KBr)
1.15 (3 H, s, 19-CH3); 13C NMR (CDCl3): ␦ 217.9 (s, C-17),
208.4 (s, C-3), 196.9 (s, C-7), 159.5 (s, C-14), 126.6 (s,
C-8), 50.8 (s, C-13), 47.8 (t, C-4), 44.1 (d, C-9), 37.9 (t,
C-2), 36.2 (s, C-5), 35.7 (t, C-16), 34.9 (t, C-1), 33.7 (s,
C-10), 33.4 (d, C-6), 28.1 (t, C-12), 27.0 (t, C-15), 25.4 (q,
C-18), 19.9 (t, C-11), 17.4 (t, C-20), 16.2 (q, C-19); Crystal
data: monoclinic P 21, a ϭ 11.4169 (14) Å, b ϭ 11.9152
(21) Å, c ϭ 12.3751 (18) Å, Z ϭ 4. Intensity data were
collected on a CAD-4 diffractometer with ␪/2␪ scan mode,
using monochromated MoK␣ radiation. Data were mea-
sured up to 2␪ of 55.0o. A total of 4230 reflections were
collected. Among them, 2618 were considered to be ob-
served ( Ͼ 2.08 ␴ (I)). Final agreement indices are Rf ϭ
0.038, Rw ϭ 0.039, GoF ϭ 1.52, based on anisotropic
refinement of all non-hydrogen atoms.; EIMS m/z [M]ϩ 312
(100), 297 (2), 284 (8), 267 (5), 255 (3), 241 (4), 227 (2),
213 (6), 199 (5), 188 (6), 173 (5), 161 (20), 148 (26), 133
(8), 115 (8), 105 (16); HREIMS m/z [M]ϩ 312.1726 (calcd.
for C20H24O3, 312.1725).
␯ max 2925, 2850, 1735, 1705, 1450, 1410, 1375, 1340,
1280, 1240, 1210, 1170, 1100, 1055, 1030, and 1015 cmϪ1;
1H-NMR ␦ (CDCl3) 5.68 (1 H, ddd, J ϭ 2.6, 2.6, 5.2 Hz),
3.13 (1 H, d, J ϭ 15.1 Hz) and 1.44 (1 H, dd, J ϭ 2.1, 15.1
Hz) (H-4’s), 2.47 (1 H, dd,. J ϭ 8.8, 19.1 Hz) and 2.12 (1
H, m) (H-16’s), 2.10 (1 H, m, H-14), 1.90 (1 H, ddd, J ϭ
2.8, 4.5, 13.2 Hz) and 1.35 (1 H, ddd, J ϭ 6.8, 11.2, 12.8
Hz) (H-12’s), 1.79 (1 H, br d, J ϭ 9.7 Hz, H-9), 1.25 (3 H,
s, H-19), 1.17 (1 H, ddd, J ϭ 4.0, 5.1, 8.0 Hz, H-6), 0.79 (1
H, ddd, 2.1, 4.0, 4.3 Hz) and 0.57 (1 H, dd, J ϭ 4.3, 8.0 Hz)
(H-20’s), 0.75 (3 H, s, H-18); 13C-NMR ␦ (CDCl3) 220.5 (s,
C-17), 210.3 (s, C-3), 132.7 (s, C-8), 122.2 (d, C-7), 49.6 (d,
C-14), 49.0 (t, C-4), 47.2 (s, C-13), 41.8 (d, C-9), 38.5 (t,
C-2), 36.2 (t, C-1), 35.7 (t, C-16), 32.7 (s, C-10), 31.1 (t,
C-12), 31.0 (s, C-5), 21.1 (t, C-11), 21.0 (t, C-15), 19.1 (t,
C-20), 17.6 (d, C-6), 16.3 (q, C-19), 13.3 (q, C-18); HRE-
IMS m/z [M]ϩ 298.1941 (calcd for C20H26O2 298.1932);
EIMS m/z [M]ϩ 298 (100), 283 (15), 270 (8), 241 (16), 227
(11), 215 (18), 202 (15), 183 (11), 175 (13), 173 (23), 159
(14), 149 (33), 145 (14), 131 (19), 117 (13), 105 (27).
13: mp 178–180°C; [␣]D25-154 (c 0.5, CHCl3); IR (KBr)
␯ max 3460 (OH), 2940, 2855, 1725, 1450, 1380, 1360,
1310, 1290, 1240, 1165 and 1070 cmϪ1; 1H-NMR ␦
(CDCl3) 5.64 (1 H, dt, J ϭ 5.0, 2.5 Hz, H-7), 3.75 (1 H, tt,
J ϭ 4.8, 11.2 Hz, H-3), 2.45 (1 H, dd, J ϭ 8.6, 14.9 Hz) and
2.11 (1 H, m) (H-16’s), 2.13 (1 H, m) and 0.93 (1 H, ddd,
J ϭ 2.1, 4.7, 12.5 Hz) (H-4’s), 1.12 (1 H, ddd, J ϭ 3.4, 5.0,
7.8 Hz, H-6), 1.04 (3 H, s, H-19), 0.76 (1 H, br dd J ϭ 3.4,
3.9 Hz) and 0.50 (1 H, dd, J ϭ 3.9, 7.8 Hz) (H-20’s), 0.75
(3 H, s, H-18); 13C-NMR ␦ (CDCl3) 220.5 (s, C-17), 69.7
(d, C-3), 132.6 (s, C-8), 122.6 (d, C-7), 49.5 (d, C-14), 42.5
(t, C-4), 47.2 (s, C-13), 41.9 (d, C-9), 31.9 (t, C-2), 34.7 (t,
C-1), 35.7 (t, C-16), 32.4 (s, C-10), 31.0 (t, C-12), 29.4 (s,
C-5), 20.9 (t, C-11), 20.7 (t, C-15), 19.5 (t, C-20), 17.9 (d,
C-6), 16.7 (q, C-19), 13.2 (q, C-18); HREIMS m/z [M]ϩ
300.2092 (calcd for C20H28O2 300.2089); EIMS m/z [M]ϩ
300 (100), 282 (22), 267 (27), 254 (9), 239 (16), 225 (14),
215 (69), 202 (27), 183 (22), 173 (31), 159 (35), 157 (50),
145 (40), 131 (47), 119 (31), 105 (72).
25
15: mp 208–210°C; [␣]D ϩ196 (c 0.3, CHCl3); IR
(KBr) ␯ max 3125, 1790, 1760, 1700, 1680, 1635, 1495,
1435, 1400, 1320, 1305, 1280, 1260, 1225, 1180, 1125,
1090, 1075 cmϪ1; UV(MeOH) ␭
(log ⑀) 245 nm (4.04);
max
1H NMR (CDCl3): ␦ 3.12 (1 H, d, J ϭ 15.3 Hz, H-4␤) and
1.62 (1 H, dd, J ϭ 15.3, 1.8 Hz, H-4␣), 2.50–2.47 (3 H, m,
H-2␤ and H-16), 2.40–2.32 (5 H, m, H-14␣, H-11 and
H-15), 2.14–2.08 (2 H, m, H-1␤ and H-2␣), 1.91–1.83 (3 H,
m, H-6␤, H-1␣ and H-12␤), 1.51 (1 H, m, H-12␣), 1.42 (3
H, s, 19-CH3), 1.01 (2 H, d, J ϭ 6.6 Hz, H-20), 0.79 (3 H,
s, 18-CH3); 13C NMR (CDCl3): ␦ 219.3 (s, C-17), 207.5 (s,
C-3), 197.3 (s, C-7), 155.1 (s, C-9), 127.8 (s, C-8), 47.1 (s,
C-13), 46.5 (t, C-4), 44.5 (d, C-14), 39.1 (s, C-10), 37.4 (t,
C-16), 37.0 (t, C-2), 33.2 (d, C-6), 32.5 (t, C-1), 30.5 (s,
C-5), 28.3 (t, C-12), 24.6 (t, C-11), 23.6 (q, C-19), 22.9 (t,
C-15), 21.8 (t, C-20), 13.3 (q, C-18); EIMS m/z [M]ϩ 312
(100), 297 (10), 284 (11), 269 (18), 256 (17), 241 (11), 227
(25), 213 15), 199 (10), 185 (10), 171 (9), 157 (8), 141 (8),
128 (10), 115 (10), 105 (8); HREIMS m/z [M]ϩ 312.1728
(calcd. for C20H24O3, 312.1725)
25
14: mp 188–189°C; [␣]D ϩ195 (c 0.5, CHCl3); IR
Additional data of 16: mp 197–198°C (lit. 194–204°C
(KBr) ␯ max 3130, 3100, 3030, 1790, 1770, 1760, 1720,
[14]); [␣]D25 ϩ273 (c 0.15, CHCl3); IR (KBr): 3100, 3020,
1705, 1645, 1490, 1410, 1340, 1310, 1290, 1250, 1200,
1790, 1715, 1700, 1640, 1490, 1400, 1290, 1245, 1065,
1170, 1095, 1065,1020 cmϪ1; UV (MeOH) ␭
(log ⑀)
max
1
1040 cmϪ1; UV (MeOH) ␭
(log ⑀) 257 nm (4.28); H
1
max
257.5 nm (3.87); H NMR (CDCl3): ␦ 3.48 (1 H, ddd, J ϭ
1.8, 10.4, 18.5 Hz, H-15␤) and 2.97 (1 H, dddd, J ϭ 3.8,
3.8, 11.4, 18.5 Hz, H-15␣), 3.13 (1 H, d, J ϭ 15.0 Hz,
H-4␤) and 1.54 (1 H, dd, 1.8, 15.0 Hz, H-4␣), 2.71 (1 H,
ddd, J ϭ 3.2, 11.3, 19.6 Hz, H-16␤) and 2.19 (1 H, ddd, J ϭ
2.0, 10.8, 19.6 Hz, H-16␣), 2.50 (2 H, m, H-2), 2.25 (1 H,
dd, J ϭ 3.0, 11.0 Hz, H-9␣), 2.06 (1 H, ddd, J ϭ 2.2, 6.5,
13.1 Hz, H-1␤), 1.87 (1 H, ddd, J ϭ 3.2, 3.2, 13.1 Hz,
H-12␤) and 1.47 (1 H, dd, J ϭ 3.2, 13.1 Hz, H-12␣),
1.79–1.75 (2 H, m, H-6␤ and H-11␤), 1.68–1.57 (2 H, m,
H-1␣ and H-11␣), 1.66 (1 H, dd, J ϭ 5.3, 11.0 Hz) and 0.88
(1 H, dd, J ϭ 5.3, 8.8 Hz) (H-20’s), 1.20 (3 H, s, 18-CH3),
NMR (CDCl3): ␦ 5.94 (1 H, s, H-4), 2.51 (1 H, ddd, J ϭ 5.6,
5.6, 13.5 Hz, H-16␤) and 2.08 (1 H, m, H-16␣), 2.44 (1 H,
m, H-2␤) and 2.37 (1 H, dd, J ϭ 4.2, 17.4 Hz, H-2␣),
2.17–2.12 (2 H, m, H-8␤ and H-12␤), 2.04 (1 H, m, H-15␤),
1.91–1.85 (2 H, m, H-1␤ and H-12␣), 1.79 (1 H, ddd, J ϭ
3.8, 13.6, 16.4 Hz, H-6␤), 1.72–1.53 (2 H, m, H-15␣ and
H-1␣), 1.42–1.33 (2 H, m, H-7␤ and H-14␣), 1.28 (1 H, m,
H-11␤) and 0.86 (1 H, m, H-11␣), 1.14 (3 H, s, 19-CH3),
0.93 (3 H, s, 18-CH3), 0.88 (1 H, dd, J ϭ 4.8, 9.3 Hz) and
0.48 (1 H, dd, J ϭ 4.8, 9.8 Hz) (H-20’s), 0.79 (1 H, ddd, J ϭ
3.6, 3.6, 15.1 Hz, H-9␣); 13C NMR (CDCl3): ␦ 220.1 (s,

J.-L. Yan / Steroids 65 (2000) 863–870
867
C-17), 198.9 (s, C-3), 171.3 (s, C-5), 126.8 (d, C-4), 49.7 (d,
C-14), 47.7 (s, C-13), 41.9 (d. C-9), 36.6 (s, C-10), 35.8 (t,
C-16), 33.94 (t, C-2), 33.86 (t, C-1), 31.33 (t, C-12), 31.24
(d, C-8), 21.5 (t, C-15), 19.6 (t, C-11), 17.1 (q, 19-CH3),
15.8 (d, C-6), 13.93 (q, 18-CH3), 13.95 (d, C-7), 8.5 (t,
C-20); Crystal data: orthorhombic P 212121, a ϭ 6.4872
(14) Å, b ϭ 8.9494 (16) Å, c ϭ 27.594 (5) Å, V Å3 ϭ
1602.0 (5), Z ϭ 4. Intensity data were collected on a CAD-4
diffractometer with ␪/2␪ scan mode, using monochromated
MoK␣ radiation. Data were measured up to 2␪ of 55.0o.
A
total of 2144 reflections were collected. Among them, 990
were considered to be observed ( Ͼ 2.0 ␴ (I)). Final agree-
ment indices are Rf ϭ 0.057, Rw ϭ 0.053, GoF ϭ 1.24,
based on anisotropic refinement of all non-hydrogen atoms.;
EIMS m/z [M]ϩ 298 (100), 283 (15), 270 (13), 256 (15),
241 (10), 227 (10), 213 (10), 199 (11), 185 (10), 172 (11),
157 (14), 149 (30), 131 (28), 117 (27), 105 (34); HREIMS
m/z [M]ϩ 298.1927 (calcd. for C20H26O2, 298.1933).
Scheme 2.
2.8.2. 6␣-Methylandrost-4-ene-3,17-dione (18)
A solution of 7 (30 mg, 0.102 mmol) and Hg(O2CCF3)2
(96 mg, 0.225 mmol) in CH2Cl2 (3 ml) was reacted as
mentioned above. The final product 18 (13.5 mg, 45%) was
obtained after chromic acid oxidation, purification and crys-
2.8. Preparation of 6-methyl steroids [15]
25
tallization. Its mp 165–167°C (lit. [16] 164–167°C); [␣]D
ϩ187 (c 0.5, CHCl3); HREIMS m/z [M]ϩ 300.2089 (calcd.
for C20H28O2, 300.2089); was identical in every respects
(IR, UV, and 1H-NMR) to 18 reported in the literature
[16,17].
2.8.1. 6␤-Methylandrost-4-ene-3,17-dione (17)
A solution of 10 (40 mg, 0.134 mmol) and Hg(O2CCF3)2
(124 mg, 0.3 mmol) in CH2Cl2 (4 ml) was stirred at r. t. for
18 h. Saturated NaCl (4 ml) was added and the two phases
were shaken vigorously for 20 min. The aqueous layer was
extracted with CH2Cl2 (6 ml x 3). The combined organic
layers were dried (Na2SO4) and concentrated in vacuo to
provide a viscous residue. This residue was dissolved in
THF (2 ml) and added to a solution of LiAlH4 (34 mg, 0.9
mmol) in THF (4 ml) at 0°C under N2. After 1 h at 0°C, the
reaction mixture was diluted with Et2O (10 ml) and
quenched sequentially with (1) H2O (0.2 ml), (2) 1 N NaOH
(0.2 ml), and (3) H2O (0.6 ml). The resulting precipitate was
stirred for 30 min, filtered, and washed with hot EtOAc. The
filtrate was concentrated in vacuo to provide a precipitate
(34 mg). This precipitate was dissolved in Me2CO (4 ml)
and cooled to 0–5°C, then 0.4 ml (0.96 mmol) of CrO3-
H2SO4 solution (prepared by added 26.8 g of CrO3 in 23 ml
of conc. H2SO4 and H2O added to make 100 ml) was added
dropwise with stirring and the reaction mixture was stirred
at r. t. for 30 min. Then i-PrOH (1 ml) and H2O (6 ml) were
added to terminate the reaction. The mixture was extracted
with CHCl3 (15 ml x 3), and the combined CHCl3 extracted
was washed with saturated NaCl solution and dried over
Na2SO4. Evaporation of the solvent in vacuo and purifica-
tion of the product over a small Si gel column (20 g) gave
analytical sample of 17 (24 mg, 60%) after recrystallization
from aceton/pet. ether. Its mp 208–209°C (lit. 207–212°C
3. Results and discussion
3.1. Biotransformation of 4
Incubation of 4 [8,9] with Mycobacterium sp. (NRRL
B-3805) gave three products, 7, 8, and 9 (Scheme 2). Com-
pound 7 and 8 were identical in every respects (mp, IR, and
1H-NMR) to 5,6␣-cyclopropano-5␣-androsta-3,17-dione
and 3␤-hydrosy-5,6␣-cyclopropano-5␣-androsta-17-one re-
ported in the literature [13], respectively. Compound 9 was
found to be identical in every respect (mmp, IR, UV, and
1H-NMR) to an authentic sample of androst-4-ene-3,17-
dione (Aldrich, Milwaukee, WI). The isolation of 9 (6%)
showed that the 5␣,6␣-cyclopropyl ring was eliminated by
this microorganism in a small extent. A total of 52.6% yield
of 17-keto steroid products were obtained.
3.2. Biotransformation of 5
Compound 5 [8,10] differs from 4 by having a 5␤,6␤-
cyclopropyl ring and an A/B cis ring junction. When it was
incubated with M. sp. (NRRL B-3805), 4%of 9 together
with two 17-keto products, 10 and 11, were isolated
(Scheme 2). Compound 10 was identical in every respect
(mp, IR, and 1H-NMR) to 5,6␤-cyclopano-5␤-androsta-
3,17-dione reported in the literature [13]. Compound 11, mp
25
[16]); [␣]D ϩ138 (c 0.5, CHCl3); HREIMS m/z [M]ϩ
300.2091 (calcd. for C20H28O2, 300.2089); was identical in
every respects (IR, UV, and 1H-NMR) to 17 reported in the
literature [16,17].
25
163–164°C, [␣]D ϩ21 (c 1.0, CHCl3), had a molecular

868
J.-L. Yan / Steroids 65 (2000) 863–870
Fig. 2. X-Ray diffraction analysis diagram of 6.
29.1 (s, C-5) in its 13C-NMR spectrum also support the
methylenation at C-5 and C-6. The structure of 6 including
the stereochemistry of the cyclopropyl group was finally
confirmed by X-ray diffraction analysis (Fig. 2).
Incubation of 6 led to the isolation of 9 (8%), 12 (10%),
13 (41%), 14 (4%), 15 (2%), and 16 (4%) (Scheme 3).
Compound 12, mp 210–213°C, [␣]D -63 (c 0.5, CHCl3),
Fig. 1. X-Ray diffraction analysis diagram of 11.
25
had a molecular formula C20H26O2 as deduced from its
HREIMS. Its IR spectrum had typical steroidal 3,17-di-
ketone absorptions at 1735 and 1705 cmϪ1 as those in 7.
The presence of this diketone function was supported by its
13C-NMR spectrum displaying two signals at ␦ 220.5 (s,
C-17) and 210.3 (s, C-3). The proton signals at ␦ 0.57 (1 H,
formula C20H30O2 as reduced from its HREIMS. Its IR
absorption at 3500 and 1725 cmϪ1 suggested the presence
of a hydroxyl and a carbonyl function, which were also
supported by 1H-NMR (␦H-3 4.03, m) and 13C-NMR spectra
(␦C-3 67.8, d; ␦C-17 221.1, s). Comparison of the 13C-NMR
of 10 and 11 revealed that 11 to be the C-3 reduced analog
of 10. Oppenauer oxidation [18] of 11 gave 10 as the sole
product was also supportive. The proton signal at ␦ 3.85 (1
H, m, H-3␣) in 5 had shifted to ␦ 4.03 (1 H, m) in 11.
Therefore, the H-3 in 11 must be ␤-oriented and the struc-
ture of 11 was established as 3␣-hydroxy-5,6␤-cyclopro-
pano-5␤-androsta-17-one. This structure assignment was
finally confirmed by X-ray diffraction analysis (Fig. 1). A total
of 57% yield of 17-ketosteroidal products were obtained.
In order to confirmed the above observation that regard-
less of their stereochemistry, the 5,6-cyclopropyl groups on
the cholestane molecule could be eliminated by this micro-
organism in small amount, compound 7 and 10 were incu-
bated with this microorganism under the same conditions,
respectively. As the result, the cyclopropyl ring eliminated
product 9 was again isolated in 6 and 4% yields, respec-
tively. In both instances 70% of the unchanged substrates
were recovered (Scheme 2).
1
dd), 0.79 (1 H, ddd), and 5.68 (1 H, ddd) in the H-NMR
spectrum indicated the preservation of the cyclopropyl ring
and the double bond. Based on those data, the structure of
5,6␣-cyclopano-5␣-androst-7-ene-3,17-dione was assigned
25
to 12. Compound 13, mp 178–180°C, [␣]D -154 (c 0.5,
CHCl3), had a molecular formula C20H28O2 as deduced
from its HREIMS. Its IR absorption at 3460 and 1725 cmϪ1
3.3. Preparation and biotransformation of 6 [11]
With curiosity to find out whether this present finding of
steroidal 5,6-cyclopropyl group elimination by this micro-
organism could be extended to some analogues of 4 and/or
5, Simmons-Smith methylenation [11] of commercially
available 7-dehydrocholesterol was conducted with CH2I2/
ZnEt2, 3␤-hydroxy-5,6␣-cyclopropano-5␣-cholest-7-ene
(6) was isolated as the sole product in 72.4% yield. Com-
25
pound 6, mp. 158–159°C, [␣]D -140 (c 0.5, CHCl3), had
a molecular formula C28H46O as deduced from its HRE-
IMS. It possessed two characteristic cyclopropyl methylene
proton signal at ␦ 0.42 and 0.73, and a vinylic proton signal
1
at ␦ 5.44 in its H-NMR spectrum. The relatively upfield
shifted carbon signals at ␦ 17.9 (d, C-6), 19.2 (t, C-28) and
Scheme 3.

J.-L. Yan / Steroids 65 (2000) 863–870
869
Fig. 4. X-Ray diffraction analysis diagram of 14.
with 18-CH3 (␦ 1.20), and the coupling between C-9 (␦
44.1) and 19-CH3 (␦ 4 1.15) indicated the presence of
⌬
8(14) double bond. Finally, the structure of 5,6␣-cyclopro-
pano-5␣-androst-8(14)-ene-3,7,17-trione for 14 was as-
signed upon confirmation by X-ray diffraction analysis
(Fig. 4).
Scheme 4.
25
Compound 15, mp 208–210°C, [␣]D ϩ196 (c 0.3,
indicated the presence of a hydroxyl and a carbonyl func-
tion, which were supported by a 13C-NMR spectrum reveal-
ing two carbon signals at ␦ 69.7 (d, C-3) and 220.5 (s,
C-17). Comparison of the 13C-NMR data of 12 and 13 (see
Experimental) suggested that 13 to be dihydro derivative of
12 at C-3 position. The Oppenauer oxidation [18] of 13 gave
12 as the sole product, and the similarity of the H-3 signals
of 13 (␦ 3.75, 1 H, tt, J ϭ 4.8, 11.2 Hz) and the starting
material 6 in their 1H-NMR spectra established the structure
of 13 to be 3␤-hydroxy-5,6␣-cyclopropano-5␣-androst-4-
CHCl3), had a molecular formula C20H24O3 as deduced
from its HREIMS. Its IR spectrum also indicated the
presence of isolated carbonyls and a conjugated carbonyl
functions at 1790, 1760, 1700 and 1680 cmϪ1. Its 13C-
NMR spectrum also showed the presence of three car-
bonyl function at ␦ 219.3 (s, C-17), 207.5 (C-3), and
197.3 (s, C-7). Also the 13C-NMR spectrum indicated the
presence of a quaternary double bond at ␦ 155.1 (s, C-9)
and 127.8 (s, C-8). Its 1H-NMR spectrum showed the
preservation of the cyclopyl ring by having the ␦ 1.01 (2
25
ene-17-one. Compound 14, mp 188–189°C, [␣]D ϩ195
H, d) signals. The UV ␭
245 nm ( log ⑀ 4.04) sup-
max
(c, 0.5, CHCl3), had a molecular formula C20H24O3 as
deduced from its HREIMS. Its IR absorption at 1790, 1770,
1720, and 1705 cmϪ1 indicated the presence of isolated
carbonyls and a conjugated carbonyl functions. The UV
ported the conjugated carbonyl function. Upon compari-
son of the UV spectra of 14, which is an exocyclic double
bond, with that of 15, the double bond in 15 must be an
endocyclic double bond in regard to C-7 carbonyl group
[19]. Furthermore, from the HMBC study (Fig. 3), it
showed the coupling between C-9 (␦ 155.1) and 19-CH3
(␦ 1.42) as well as coupling between C-14 (␦ 44.5) and
18-CH3 (␦ 4 0.79) also supported the presence of a ⌬8(9)
double bond rather than the alternative ⌬8(14) double
bond. Therefore, the structure of 5,6␣-cyclopropano-5␣-
androst-8-ene-3,7,17-trione was assigned to 15.
␭
257.5 nm ( log ⑀ 3.87) supported this presence of a
max
conjugated carbonyl function. Its 13C-NMR spectrum indi-
cated the presence of three carbonyl groups at ␦ 217.9 (s,
C-17), 208.4 (s, C-3), and 196.9 (s, C-7). Also the ␦ 159.5
(s, C-14) and 126.6 (s, C-8) signals at 13C-NMR spectrum
indicated the presence of a quaternary double bond. The
1H-NMR signals at ␦ 0.88 (1 H, dd, H-20) and 1.66 (1 H,
dd, H-20) indicated the preservance of the cyclopropyl ring
in the molecule. From the HMBC study (Fig. 3), it showed
the coupling between C-14 (␦ 159.5) and C-17 (␦ 217.9)
25
Compound 16, mp 197–198°C, [␣]D ϩ273 (c 0.15,
CHCl3), had a molecular formula C20H26O2 as deduced
from its HREIMS. Its IR absorption at 1790, 1715, and 1700
cmϪ1 and UV ␭
257.0 nm ( log ⑀ 4.28) indicated the
max
presence of a conjugated carbonyl function and an isolated
carbonyl group. Its 13C-NMR showed the presence of car-
bonyl functions at ␦ 220.1 (s, C-17) and ␦ 198.9 (s, C-3).
The 1H-NMR signals at ␦ 0.48 (1 H, dd) and 0.88 (1 H, dd)
indicated the presence of a cyclopropyl ring, ␦ 5.94 (1 H, s,
H-4) indicated having the ⌬4-3-ketone system. Its mp sim-
ilar to 6␣,7␣-cyclopropanoandrost-4-ene-3,17-dione re-
ported in the literature (194–204°C [14]). This structural
assignment was finally confirmed by X-ray diffraction anal-
ysis (Fig. 5).
Fig. 3. Major HMBC correlation of 14 and 15.
It is interesting to notice that the presence of an addi-

870
J.-L. Yan / Steroids 65 (2000) 863–870
steroids. This combination of chemical and microbial
approach offers an alternative route to 6-methyl steroids
from sterols such as cholesterol, sitosterol and campes-
terol.
Acknowledgments
The authors are grateful to NSC, R.O.C. for financial
support of this research under Grants NSC 86–2314-B-002–
083 and NSC 87–2314-B-002–011.
Fig. 5. X-Ray diffraction analysis diagram of 16.
References
tional ⌬7 double bond in 6 led to the isolation of three more
metabolites than 4 and also the yield of 9, the cyclopropyl
ring eliminated product, was elevated a little. Direct incu-
bation of 12 with the same microorganism under the same
condition, the cyclopropyl ring migrated metabolite 16 was
isolated in up to 10% yield. One possible explanation for
this migration might be that when C-5-C-20 bond was
cleaved either by enzymatic reaction or due to ⌬4 introduc-
tion by ⌬4-dehydrogenase and before it was cleaved at
C-6-C-20 bond to form 9, and due to presence of the
neighboring ⌬7, it closed to form the 6␣,7␣-cyclopropyl
ring and gave 16 as the isolable product.
[1] Martin CKA. Sterols. In: Kieslich K, editor. Biotechnology, A Com-
prehensive Treatise in 8 Volumes (ed. Rehm HI, Reed G), Vol. Vla:
Biotransformations. Weinheim: Verlag-Chemie, 1984. p. 80.
[2] Kieslich K. Microbial side-chain degradion of sterols. J Basic Micro-
biol 1985;25:461–74.
[3] Marsheck WJ, Krachy S, Muir RD. Microbial degradation of sterols.
Appl Microbiol 1972;23:72–7.
[4] Sih CJ, Wang KC, Tai HH. C22 Acid intermediates in the microbio-
logical cleavage of the cholesterol side chain. J Am Chem Soc
1967;89:1956–7.
[5] Fujimoto Y, Chen CS, Gopalan AG, Sih CJ. Microbial degradation of
the phytosterol side-chain. II. Incorporation of NaH14CO3 onto C-28-
positon. J Am Chem Soc 1982;103:4720–2.
[6] Wang KC, Young LH, Lee SS. Microbial transformation of cyclo-
artenol and 24-methylene cyclartanol. Tetrahedron Lett
1990;31:1283–6.
[7] Lee SS, Young LH, Wang KC. Microbial transformation of cyclo-
artenol, 24-methylenecycloartanol, and lanosterol, II. Isolation and
characterization of C19 steroids. J Nat Prod 1991;54:178–83.
[8] Kohout L, Fajkos J. On steroids. CLI. B-Homosteroids. VIII. Solvolytic
reaction of the epimeric 3-methanesulphonyloxy-5,7-cyclo-B-homos-
teroids. Collection Czechoslov Chem Commun 1973;38:913–28.
[9] Templeton JF, Wie CW. Mass spectra of some 5,6␣-cyclopropano-
5␣-cholestane derivatives. Can J Chem 1971;49:3636–41.
[10] Kohout L, Fajko4 s J. On steroids. CXLIII. B-Homosteroids. VI.
Simmons-Smith methylenation of ⌬5 steroids. Collection Czechoslov
Chem Commun 1972;37:3490–500.
3.4. Preparation of 6-methyl steroids [15]
The 6-methyl androst-4-ene-3,17-diones are potent
competitive human placental aromatase inhibitors. The
aromatase inhibitors appear to be a major clinical role for
methods of controlling estrogen level and could be useful
in the treatment of estrogen-dependent breast cancer [17].
The previous work on the stereospecific oxymercuration
of cyclopropanes by Collum et al. [15] prompted us to
study the reaction of mercuric trifluoroacetate on these
two major microbial transformation products isolated in
this study.
Treatment of 7 and 10 with mercuric trifluoroac-
etate, followed by reductive workup with lithium alumi-
num hydride, and final oxidation of the resulting alcohols
with Jone’s reagents gave 18 and 17 as the sole products
in 45 and 60% yields, respectively. The rather lower yield
of 18 obtained could possibly be elevated to a level
similar to 17 by further optimization of reaction
conditions.
[11] Arai I, Mori A, Yamamoto H. An asymmetric Simmons-Smith reac-
tion. J Am Chem Soc 1985;107:8254–6.
[12] Breitmaier E, Voelter W. In: Carbon-13 NMR Spectroscopy. New
York: VCH, 1990. p. 349.
[13] Kohout L, Fajkos J. On steroids. CLIV. B-Homosteroids. IX. The
epimeric 5,7-cyclo-B-homoandrostane derivatives. Collection
Czechoslov Chem Commun 1973;38:1415–27.
[14] Christiansen RG, Schodack NY, Johnson WS. 7-Cyanosteroids and
derivatives thereof. US Patent 3,200,113 1965 Aug 10.
[15] Collum DB, Still WC, Mohamadi F. Synthesis of alternating hy-
droxy- and methyl-substituted hydrocarbons by oxymercuration of
cyclopropylcarbinols. J Am Chem Soc 1986;108:2094–6.
[16] Campbell JA, Babcock JC, Hogg JA. 6-Methyl steroids in the andro-
stane series. J Am Chem Soc 1958;80:4717–21.
In summary, we have described the microbial trans-
formations of 5,6-cyclopropanocholestanes with M. sp.
(NRRL B-3805). The cleavage of the cyclopropyl ring
was observed in low extents, and the major products were
17-keto steroids with intact 5,6-cyclopropyl ring. The
chemical opening of the cyclopropyl ring of these two
major products and further modifications led to 6-methyl
[17] Numazawa M, Oshibe M. 6-Alkyl- and 6-arylandrost-4-ene-3,17-
diones as aromatase inhibitors. Synthesis and structure activity rela-
tionships. J Med Chem 1994;37:1312–9.
[18] Oppenauer RV. Cholestenone. Org Synth Collective Volume 1955;
3:207–9. 4
[19] House HO, Thompson HW. Synthesis of ⌬8,9-4 and ⌬9,10-octal-1-
one. J Org Chem 1961;26:3729–34.