Transformation of a series of saturated isomeric steroidal diols by Aspergillus tamarii KITA reveals a precise stereochemical requirement for entrance into the lactonization pathway

Article abstract of DOI:10.1016/j.jsbmb.2010.08.010

Four isomers of 5α-androstan-3,17-diol have been transformed by the filamentous fungus Aspergillus tamarii, an organism which has the ability to convert progesterone to testololactone in high yield through an endogenous four step enzymatic pathway. The only diol handled within the lactonization pathway was 5α-androstan-3α,17β-diol which, uniquely underwent oxidation of the 17β-alcohol to the 17-ketone prior to its Baeyer-Villiger oxidation and the subsequent production of 3α-hydroxy-17a-oxa-D-homo-5α-androstan-17-one. This demonstrated highly specific stereochemical requirements of the 17β-hydroxysteroid dehydrogenase for oxidation of this specific steroidal diol to occur. In contrast, the other three diols were transformed within the hydroxylation pathway resulting in functionalization at C-11β. Only 5α-androstan-3β,17α-diol could bind to the hydroxylase in multiple binding modes undergoing monohydroxylation in 6β and 7β positions. Evidence from this study has indicated that hydroxylation of saturated steroidal lactones may occur following binding of ring-D in its open form in which an α-alcohol is generated with close spatial parity to the C-17α hydroxyl position. All metabolites were isolated by column chromatography and were identified by ¹H, ¹³C NMR and DEPT analysis and further characterized using infra-red, elemental analysis and accurate mass measurement.

Full text of DOI:10.1016/j.jsbmb.2010.08.010

Journal of Steroid Biochemistry & Molecular Biology 122 (2010) 352–358
Contents lists available at ScienceDirect
Journal of Steroid Biochemistry and Molecular Biology
journal homepage: www.elsevier.com/locate/jsbmb
Transformation of a series of saturated isomeric steroidal diols by Aspergillus
tamarii KITA reveals a precise stereochemical requirement for entrance
into the lactonization pathway
A. Christy Hunter∗, Catherine Collins, Howard T. Dodd, Cinzia Dedi, Salomé-Juliette Koussoroplis
Molecular Targeting and Polymer Toxicology Group, School of Pharmacy, The Huxley Biosciences Building, University of Brighton, East Sussex BN2 4GJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 July 2010
Received in revised form 12 August 2010
Accepted 31 August 2010
Four isomers of 5␣-androstan-3,17-diol have been transformed by the filamentous fungus Aspergillus
tamarii, an organism which has the ability to convert progesterone to testololactone in high yield through
an endogenous four step enzymatic pathway. The only diol handled within the lactonization path-
way was 5␣-androstan-3␣,17␤-diol which, uniquely underwent oxidation of the 17␤-alcohol to the
17-ketone prior to its Baeyer–Villiger oxidation and the subsequent production of 3␣-hydroxy-17a-oxa-
D-homo-5␣-androstan-17-one. This demonstrated highly specific stereochemical requirements of the
17␤-hydroxysteroid dehydrogenase for oxidation of this specific steroidal diol to occur. In contrast,
the other three diols were transformed within the hydroxylation pathway resulting in functionaliza-
tion at C-11␤. Only 5␣-androstan-3␤,17␣-diol could bind to the hydroxylase in multiple binding modes
undergoing monohydroxylation in 6␤ and 7␤ positions. Evidence from this study has indicated that
hydroxylation of saturated steroidal lactones may occur following binding of ring-D in its open form in
which an ␣-alcohol is generated with close spatial parity to the C-17␣ hydroxyl position. All metabolites
were isolated by column chromatography and were identified by 1H, 13C NMR and DEPT analysis and
further characterized using infra-red, elemental analysis and accurate mass measurement.
© 2010 Elsevier Ltd. All rights reserved.
Keywords:
Aspergillus tamarii
Steroidal diol
Fungus
17␤-Hydroxysteroid dehydrogenase
Baeyer–Villiger oxidase
Hydroxylation
Lactonization
1. Introduction
cell cultures of A. tamarii. These studies have demonstrated that a
wide-range of structurally diverse steroids can be handled by the
The fungus Aspergillus tamarii contains an endogenous sequen-
tial enzymatic pathway (Fig. 1) which converts progesterone (1) via
Baeyer–Villiger oxidation to testosterone acetate (2). This under-
goes hydrolysis to form testosterone (3) and is then oxidised
to androst-4-en-3,17-dione (4). A final Baeyer–Villiger oxidation
of the C-17 ketone affords testololactone (5) in high yield [1].
Divergence from this pathway can occur through testosterone
(3) which, can be functionalized by a hydroxylase affording 11␤-
hydroxytestosterone. Progesterone (1) can also undergo reversible
reduction of the C-20 ketone resulting in (6), the equilibrium
between the reductase and the oxidase is thought to be dynamic
[2]. 20(R)-Reduction products of progesterone analogues have been
isolated in a number of studies [2,3].
endogenous pathways within this organism, and have highlighted
the powerful effect of single functional groups in directing their
metabolic fate (oxidative or reductive) [3,6,7]. To date, no single
study has determined the metabolic fate of all four isomers of 5␣-
androstan-3,17-diol. This is of significance due to the importance
of hydrogen bonding in steroid/enzyme interactions [8–16] and in
determining if specific stereochemistry is handled within differ-
ent metabolic pathways. We have also modified a previous [17]
synthetic strategy to facilitate an expedient route to 3␣-hydroxy
containing 17␣- and 17␤-diols that, as with the other analogues
[18–23] have a broad range of interesting biological activity in their
own right [24–27].
Steroids differing in side-chain structure [2,3] substituent posi-
tion [4,5] and ring-Dstructure [6,7] have been transformed in whole
2. Materials and methods
2.1. Chemicals and reagents
5␣-Androstan-3␤,17␣-diol (13) and 5␣-androstan-3␤,17␤-diol
(14) were synthesized as previously described [17] and were of high
purity (99%+ by carbon content) as determined by elemental analy-
sis found: C, 77.39; H, 11.47. C19H32O2 requires C, 78.03; H, 11.03%
∗
Corresponding author at: Molecular Targeting and Polymer Toxicology Group,
School of Pharmacy, The Huxley Biosciences Building, University of Brighton, East
Sussex BN2 4GJ, UK. Tel.: +44 01273 642088; fax: +44 01273 642674.
E-mail address: c.hunter@bton.ac.uk (A.C. Hunter).
0960-0760/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsbmb.2010.08.010

A.C. Hunter et al. / Journal of Steroid Biochemistry & Molecular Biology 122 (2010) 352–358
353
Fig. 1. The endogenous lactonization pathway present in Aspergillus tamarii KITA, that converts progesterone (1) to testololactone (5) with the preceding C-20 keto-alcohol
reducto-oxidase interconverting (1) and (6) and the minor 11␤-hydroxylation pathway from testosterone (3).
Table 2
for compound (13) and found: C, 77.50; H, 11.54. C19H32O2 requires
13C NMR data for synthesis of the 3␣-containing diols determined in CDCl3.
C, 78.03; H, 11.03% for compound (14). 3␤-Hydroxy-5␣-androstan-
17-one was the starting material for the synthetic sequence to
compounds (13) and (14) was purchased from Steraloids Ltd. (UK).
From the same supplier 3␣-hydroxy-5␣-androstan-17-one was
obtained and initiated the reaction sequence to 5␣-androstan-
3␣,17␣-diol (11) and 5␣-androstan-3␣,17␤-diol (15) which were
also found to be of high purity (99%+ by carbon content) as deter-
mined by elemental analysis found: C, 77.75; H, 11.77. C19H32O2
requires C, 78.03; H, 11.03% for compound (11) and found: C, 77.76;
H, 10.85. C19H32O2 requires C, 78.03; H, 11.03% for (15). Solvents
were of analytical grade; petroleum ether refers to the fraction
with a boiling point of 60–80 ◦C. Silica for column chromatogra-
phy was Merck 9385 and TLC was performed with Macherey-Nagel
Carbon atom
Compounds
7
8
9
10
32.05
11
12
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
31.56
30.85
66.37
35.86
35.04
28.25
29.00
39.11
51.50
35.47
21.75
32.14
47.48
54.43
20.04
35.78
31.48
29.04
66.49
35.83
39.16
24.35
28.39
34.93
53.75
36.28
20.43
34.28
44.07
54.68
23.28
32.16
163.90
17.06
11.18
32.01
31.48
66.51
35.81
39.31
28.37
29.00
34.57
54.67
36.30
20.63
36.26
50.04
54.79
33.63
32.32
29.03
66.55
35.91
39.10
28.55
32.26
35.76
54.13
36.40
24.60
32.32
45.53
48.78
20.28
31.49
80.06
17.11
11.20
32.22
29.02
66.57
37.24
39.12
28.50
32.13
35.33
54.47
36.23
20.40
38.74
41.89
52.14
35.89
71.87
52.04
18.70
11.19
31.97
66.58
35.94
39.36
28.56
29.04
34.12
55.17
45.39
20.75
37.16
45.58
56.17
32.09
129.29
144.00
17.08
11.18
Alugram® SIL G/UV254
.
2.2. Formation of hydrazone (8)
112.93
137.48
15.28
221.10
11.18
13.83
A solution of 3␣-hydroxy-5␣-androstan-17-one (7) (3 g) in
ethanol (30 cm3) was treated with triethylamine (6 cm3) and
hydrazine hydrate (20 cm3) and the solution was heated under
reflux for 3 h. The solution was poured into cold water to give
3␣-hydroxy-5␣-androstan-17-hydrazone (8) (3.1 g) which crystal-
lized from ethanol as cubes m.p. 239–241 ◦C found: C, 74.79; H,
10.37; N, 9.03. C19H32N2O requires C, 74.95; H, 10.59; N, 9.20%. For
spectroscopic data please refer to Tables 1–3.
11.14
Table 3
Significant infra-red absorption signals for compounds 7–12.
Compound
–OH
>C
O
>C C<
3␣-Hydroxy-5␣-androstan-17-one (7)
3␣-Hydroxy-5␣-androstan-17-hydrazone (8)
17-Iodo-5␣-androst-16-en-3␣-ol (9)
5␣-Androst-16-en-3␣-ol (10)
5␣-Androstan-3␣,17␣-diol (11)
5␣-Androstan-3␣,16␣-diol (12)
3379
3343
3343
3276
3344
3261
1724
1628
1588
2.3. Formation of vinyl iodide (9)
3␣-Hydroxy-5␣-androstan-17-hydrazone (8) (3 g) in dry THF
(70 cm3) and triethylamine (30 cm3) under nitrogen was treated
with iodine (4 g in 12.5 cm3 THF) at room temperature. After nitro-
gen evolution had ceased the THF was removed in vacuo and the
mixture was extracted with ethyl acetate. The extract was washed
with dilute hydrochloric acid, water, aqueous sodium hydrogen
carbonate, saturated sodium chloride and dried over sodium sulfate
and isolated as a gum (3 g), found: 423.115 C19H29INaO requires
423.115. For spectroscopic data please refer to Tables 1–3.
2.4. Microorganism
A. tamarii KITA (QM 1223) was purchased from the collection
at CABI Bioscience (UK). Stock cultures were maintained at 4 ◦C on
potato dextrose agar slopes. Steroid transformation studies were
carried out in 3% malt extract medium (Oxoid, UK).
Table 1
1H NMR data for synthesis of the 3␣-containing diols determined in CDCl3.
Compound
3␤-H
18-H3
19-H3
Other significant signals
3␣-Hydroxy-5␣-androstan-17-one (7)
3␣-Hydroxy-5␣-androstan-17-hydrazone (8)
17-Iodo-5␣-androst-16-en-3␣-ol (9)
5␣-Androst-16-en-3␣-ol (10)
5␣-Androstan-3␣,17␣-diol (11)
5␣-Androstan-3␣,16␣-diol (12)
4.05 (t, J = 2.6 Hz)
4.04 (t, J = 2.6 Hz)
4.05 (t, J = 2.6 Hz)
4.04 (t, J = 2.6 Hz)
4.04 (t, J = 2.6 Hz)
4.04 (t, J = 2.6 Hz)
0.86
0.84
0.72
0.75
0.65
0.70
0.80
0.80
0.80
0.81
0.79
0.77
4.73 (2H, brs, NH2)
6.11 (1H, dd, J = 2 Hz, J = 3.5 Hz, 16-H)
5.69 (1H, brs, 17-H) 5.83 (1H, dd, J = 1.4 Hz, J = 4.3 Hz, 16-H)
3.71 (1H, d, J = 6 Hz, 17␤-H)
4.44 (1H, m, ␻/2 = 10 Hz, J = 6.5 Hz, 16␤-H)

354
A.C. Hunter et al. / Journal of Steroid Biochemistry & Molecular Biology 122 (2010) 352–358
Table 4
1H NMR data for steroidal starting material (except compound 11, see Table 1) and transformation products determined in CDCl3.
Reference data
3-H
18-H3
19-H3
Other significant signals
5␣-Androstan-3␤,17␣-diol (13)
5␣-Androstan-3␤,17␤-diol (14)
5␣-Androstan-3␣,17␤-diol (15)
3.60 (1H, tt, J = 5 Hz, J = 10 Hz, 3␣-H)
0.65
0.73
0.73
0.82
0.81
0.79
3.73 (1H, d, J = 5 Hz, 17␤-H)
3.60 (2H, m, ␻/2 = 16 Hz, 3␣-H, 17␣-H)
3.63 (1H, t, J = 8 Hz, 17␣-H)
4.04 (1H, t, J = 2.6 Hz, 3␤-H)
Transformation products
5␣-Androstan-3␣,11␤,17␣-triol (16)
5␣-Androstan-3␤,6␤,17␣-triol (17)
5␣-Androstan-3␤,7␤,17␣-triol (18)
4.05 (1H, t, J = 2.3 Hz, 3␤-H)
3.60 (1H, tt, J = 5 Hz, J = 10 Hz, 3␣-H)
3.60 (1H, tt, J = 5 Hz, J = 10 Hz, 3␣-H)
0.90
0.89
0.68
1.00
1.26
0.85
3.70 (1H, d, J = 6 Hz, 17␤-H) 4.61 (1H, s, 11␣-H)
3.70 (1H, d, J = 6 Hz, 17␤-H) 4.60 (1H, s, 6␣-H)
3.40 (1H, td, J = 5 Hz, J = 10 Hz, 7␣-H) 3.70 (1H, d,
J = 6 Hz, 17␤-H)
5␣-Androstan-3␤,11␤,17␣-triol (19)
5␣-Androstan-3␤,11␤,17␤-triol (20)
3.60 (1H, tt, J = 5 Hz, J = 10 Hz, 3␣-H)
4.06 (1H, t, J = 2.6 Hz, 3␤-H)
0.89
0.98
1.00
1.05
3.70 (1H, d, J = 6 Hz, 17␤-H) 4.61 (1H, s, 11␣-H)
3.59 (2H, m, ␻/2 = 16 Hz, 3␣-H, 17␣-H) 4.33 (1H,
s, 11␣-H)
3␣-Hydroxy-17a-oxa-D-homo-5␣-
0.75
1.30
androstan-17-one (21)
Table 6
2.5. Conditions of cultivation and transformation
Significant metabolite infra-red absorption signals.
Spores were transferred aseptically in a category 2 biological
safety cabinet into 500 ml Erlenmeyer flasks containing 300 ml
of sterile media and were incubated for 72 h at 30 ◦C. The cul-
tures were shaken at 180 rpm on an orbital shaker. Aliquots (5 ml)
from the seed flask were transferred aseptically to 10 flasks and
grown for a further 72 h as above, at the end of which the fun-
gus is in log phase growth. After this time period steroid dissolved
in dimethylformamide (DMF) was evenly distributed between the
flasks (1 mg/ml) under sterile conditions and incubated for a fur-
ther 5 days after which, the metabolites were extracted from the
broth.
Compound
–OH
Lactone
5␣-Androstan-3␤,17␣-diol (13)
5␣-Androstan-3␤,17␤-diol (14)
5␣-Androstan-3␣,17␤-diol (15)
5␣-Androstan-3␣,11␤,17␣-triol (16)
5␣-Androstan-3␤,6␤,17␣-triol (17)
5␣-Androstan-3␤,7␤,17␣-triol (18)
5␣-Androstan-3␤,11␤,17␣-triol (19)
5␣-Androstan-3␤,11␤,17␤-triol (20)
3␣-Hydroxy-17a-oxa-D-homo-5␣-
androstan-17-one (21)
3253 br
3456, 3196
3322 br
3366 br
3492, 3320
3358 br
3358 br
3381 br
3534 br
1712
acetate or pure ethyl acetate. A 50:50 sulfuric acid in methanol
spray was used to develop the TLC plates.
2.6. Extraction and identification of metabolites
The fungal mycelium was separated from the broth by filtra-
tion under vacuum. Following completion the mycelium was rinsed
with ethyl acetate (0.5 L) to ensure the entire available steroid was
removed. The mycelial broth was then extracted thrice with ethyl
acetate (1.5 L). The organic extract was dried over sodium sulfate
and the solvent evaporated in vacuo to give a gum. This gum was
adsorbed onto silica and chromatographed on a column of silica;
the steroidal metabolites were eluted with increasing concentra-
tions of ethyl acetate in petroleum ether. The solvent was collected
in aliquots (10 ml) and analysed by thin layer chromatography (TLC)
to identify the separated metabolite fractions. The solvent systems
used for running the TLC plates were 50:50 petroleum ether in ethyl
2.7. Analysis and identification of metabolites
Characteristic splitting patterns [28] and shift values [29]
in the 1H and 13C NMR spectra in comparison to the starting
compounds were used to determine metabolite structure in com-
bination with DEPT analysis to identify the nature of the carbon
(Tables 1, 2, 4 and 5). Spectra were recorded on a Bruker WM 360
Spectrometer, all samples were analysed in deuteriochloroform
using tetramethylsilane as the internal standard. High resolution
mass measurement (HRMS) was determined in electrospray ion-
ization (ESI) mode using a Bruker Daltonics Microtof spectrometer.
Infra-red absorption spectra (Tables 3 and 6) were recorded directly
Table 5
13C NMR data for starting materials (13–15) and biotransformation products (16–21) determined in CDCl3.
Carbon atom
Compounds
13
14
15
16
17
18
19
20
21
37.49
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
32.40
31.48
71.36
38.11
44.80
28.68
32.28
35.75
54.48
37.73
20.75
37.08
43.02
48.74
24.61
31.62
80.08
17.09
12.35
36.75
30.54
71.31
38.18
44.91
28.57
31.63
35.55
54.47
36.81
20.84
37.03
42.49
51.01
23.40
31.51
81.99
11.15
12.36
32.22
29.02
66.56
36.76
39.20
28.43
31.57
35.55
54.47
36.20
20.36
35.88
42.99
51.09
23.36
30.52
81.99
11.15
11.21
32.71
28.65
68.56
35.31
39.89
27.99
32.48
31.30
57.98
30.97
66.43
40.42
44.00
50.38
24.54
37.00
79.51
19.43
14.35
37.64
31.14
68.87
32.73
50.32
71.05
40.41
31.33
58.03
35.38
20.58
36.98
42.74
45.80
24.55
32.56
79.44
17.16
15.49
32.68
31.35
71.06
36.96
42.07
37.69
75.44
43.59
52.26
35.08
20.86
38.17
46.18
47.84
27.22
31.42
78.88
17.18
12.47
37.00
31.20
71.09
37.70
45.74
28.16
32.73
31.35
58.04
35.71
68.91
40.43
43.34
50.33
24.56
32.54
79.49
19.41
15.52
36.89
30.35
72.23
37.69
45.80
28.05
31.96
31.33
58.18
35.70
71.13
45.84
42.05
52.51
23.45
31.17
82.42
13.61
15.46
31.95
66.30
35.57
39.33
28.13
30.53
36.09
53.70
35.57
28.68
30.49
83.44
46.35
19.76
28.68
172.50
20.14
11.07

A.C. Hunter et al. / Journal of Steroid Biochemistry & Molecular Biology 122 (2010) 352–358
355
Table 7
Product yields following chromatography.
Starting material
Metabolites
Yield (%)
5␣-Androstan-3␣,17␣-diol (11)
6
12
8
5␣-Androstan-3␣,11␤,17␣-triol (16)
5␣-Androstan-3␤,17␣-diol (13)
5␣-Androstan-3␤,6␤,17␣-triol (17)
5␣-Androstan-3␤,7␤,17␣-triol (18)
5␣-Androstan-3␤,11␤,17␣-triol (19)
3
55
8
5␣-Androstan-3␤,17␤-diol (14)
5␣-Androstan-3␣,17␤-diol (15)
1
1
3
5␣-Androstan-3␤,11␤,17␤-triol (20)
3␣-Hydroxy-5␣-androstan-17-one (7)
3␣-Hydroxy-17a-oxa-D-homo-5␣-androstan-17-one (21)
23
29
on a Nicolet avatar 320 FT-IR fitted with a Smart Golden Gate®. All
yields obtained from the transformation experiments are listed in
Table 7.
3␤,6␤,17␣-triol (17) the 19-methyl protons had undergone a
significant downfield shift (0.44 ppm) in comparison to the starting
material (13). The product spectrum also contained a new reso-
nance signal at 4.61 ppm (1H, s) indicating hydroxylation at an axial
proton, possibly 6␤-H. This notion was confirmed by comparison of
the 13C NMR spectrum of (17) with that of (13). The beta carbons (C-
5 and C-7) had undergone downfield shifts (5.52 ppm and 8.13 ppm
respectively) as had C-19 (3.14 ppm) and carbons gamma to C-6 had
undergone upfield shifts C-4 (5.38 ppm), C-8 (4.42 ppm) and C-10
(2.35 ppm). Monohydroxylation was confirmed by accurate mass
measurement(HRMSESI: calc. forM+Na331.224C19H32NaO3 obsd.
331.223).
3. Results
3.1. Confirmation of structure for the novel synthetically derived
compounds (8) and (9)
Comparison of the 1H NMR spectrum of (8) with the starting
material (7) demonstrated the presence of a new signal at 4.73 ppm
(2H, brs) which was consistent with the amino protons (–NH2).
Examination of the 13C NMR spectrum revealed a range of signal
shifts in the product consistent with replacement of the more elec-
tronegative C-17 carbonyl by the hydrazone. These included upfield
shifts in C-16 (12.5 ppm), C-17 (57.2 ppm), C-13 (3.41) and a down-
field shift for C-18 (5.88 ppm). The structure was fully supported by
elemental analysis (Section 2.2) and accurate mass measurement
(HRMS ESI: calc. for M+Na 327.240 C19H32N2NaO obsd. 327.240).
Comparison of the 1H and 13C NMR spectra of the vinyl iodide
(9) with that of the hydrozone (8) enabled confirmation of struc-
ture. The product spectrum (9) was devoid of the resonance signal
at 4.73 (2H, brs) consistent with the loss of the –NH2 and con-
tained a new signal at 6.11 ppm (1H, dd) synonymous with the
presence of a ring-D double bond. This was further confirmed in
the 13C NMR spectrum with the presence of resonance signals for
C-17 (137.48 ppm) and C-16 (112.93 ppm) consistent with a dou-
ble bond and a concomitant downfield shift for C-15 to 33.63 ppm
(ꢀ10.35 ppm). The presence of the iodine was fully supported by
accurate mass measurement (Section 2.2).
Comparison of the 1H NMR of 5␣-androstan-3␤,7␤,17␣-triol
(18) with that of (13) revealed a new signal at 3.40 ppm (1H, td)
indicating hydroxylation at an equatorial proton with no concomi-
tant shift in either of the methyl signals suggesting substitution at
C-7␤. This was supported by comparison of the 13C NMR spectrum
of this product and that of (13). Carbons beta to C-7 underwent
downfield shifts C-6 (9.01 ppm) and C-8 (7.84 ppm) with those ␥-
carbon signals shifting C-5 (2.73 ppm), C-9 (2.22 ppm) and C-14
(0.9 ppm). Accurate mass measurement confirmed that this com-
pound was monohydroxylated (HRMS ESI: calc. for M+Na 331.224
C
19H32NaO3 obsd. 331.224).
The 1H NMR spectrum of 5␣-androstan-3␤,11␤,17␣-triol (19)
contained a new signal at 4.61 ppm (1H, s) indicating axial substitu-
3.2. Products of metabolism and structural identification
Transformation of 5␣-androstan-3␣,17␣-diol (11) resulted in
the isolation of one product. The structure of 5␣-androstan-
3␣,11␤,17␣-triol (16) was determined by comparison of its’ spectra
with that of (11). The 1H NMR revealed significant downfield shifts
for both 18-H and 19-H methyl protons respectively of 0.25 ppm
and 0.21 ppm and taken into account with a new signal at 4.60 ppm
(1H, s) this strongly suggest substitution at the 11␤-position. This
was confirmed by the product 13C NMR spectra which demon-
strated ␣-carbon downfield shift for C-11 (41.83 ppm), ␤-carbon
downfield shifts for C-9 (3.85 ppm) and C-12 (8.19 ppm) and ␥-
carbon upfield shifts for C-8 (4.43 ppm), C-10 (5.43 ppm) and C-13
(1.53 ppm). Comparatively both methyl groups were also shifted
downfield C-18 (2.32 ppm) and C-19 (3.15 ppm) which is consistent
with the presence of a C-11␤ hydroxyl group. The monohydroxy-
lation was fully supported by accurate mass measurement (HRMS
ESI: calc. for M+Na 331.224 C19H32NaO3 obsd. 331.223) (Fig. 2).
Transformation of 5␣-androstan-3␤,17␣-diol (13) afforded
three products. In the 1H NMR spectrum for 5␣-androstan-
Fig. 2. Synthetic reaction pathway to 5␣-androstan-3␣,17␣-diol (11): (i)
H2NNH2/(CH3CH2)3N/EtOH; (ii) I/(CH3CH2)3N/THF; (iii) Na/EtOH; (iv) (a) B2H6/THF
and (b) H2O2/NaOH.

356
A.C. Hunter et al. / Journal of Steroid Biochemistry & Molecular Biology 122 (2010) 352–358
Fig. 3. Transformation of steroidal diols following 5 days incubation with Aspergillus tamarii KITA.
tion. In comparison to the spectrum of (13) both methyl groups had
undergone significant downfield shifts 18-H (0.24 ppm) and 19-H
(0.18 ppm) strongly suggesting hydroxylation at the 11␤-position.
This was confirmed following comparison of the product 13C NMR
spectrum to that of (13) which demonstrated ␤-carbon downfield
shifts for C-9 (3.56 ppm) and C-12 (3.35 ppm) and ␥-carbon upfield
shifts for C-8 (4.40 ppm) and C-10 (2.02 ppm). Monohydroxylation
was confirmed by accurate mass measurement (HRMS ESI: calc. for
M+Na 331.224 C19H32NaO3 obsd. 331.223).
0.44–1.30 ppm which is fully indicative of hetero-atom insertion
[2,3,6] in ring-D. This was supported in the 13C NMR spectrum
with downfield shifts for C-13 (35.96 ppm) and C-18 (8.96 ppm),
directly associated with insertion of the hetero-atom, and upfield
shifts consistent with the reduced 6 membered ring strain for C-
14 (8.08 ppm) and C-17 (48.60 ppm). Comparison of the product
spectrum to that of an authentic sample [6] fully supported the
proposed structure.
Transformation of 5␣-androstan-3␤,17␤-diol (14) afforded one
product, 5␣-androstan-3␤,11␤,17␤-triol (20). This was identified
by inspection of its’ spectra with the 1H NMR, compared to (14),
which, contained a new signal at 4.33 ppm (1H, s) suggestive of axial
substitution. This information combined with the downfield shifts
observed for 18-H (0.25 ppm) and 19-H (0.26 ppm) was supportive
of C-11␤ hydroxylation. The 13C NMR spectrum of (20) was consis-
tent with beta carbon downfield shift for C-9 (3.71 ppm) and C-12
(37.03 ppm) and ␥-carbon upfield shifts for C-8 (4.22 ppm) and C-10
(1.11 ppm). Monohydroxylation was confirmed by accurate mass
determination (HRMS ESI: calc. for M+Na 331.224 C19H32NaO3
obsd. 331.224).
4. Discussion
Transformation of this series of the four isomers of androstan-
3,17-diol revealedexclusive handlingof5␣-androstan-3␣,17␤-diol
(15) within the lactonization pathway, ultimately resulting in
the generation of 3␣-hydroxy-17a-oxa-D-homo-5␣-androstan-
17-one (21). Before the lactone (21) is formed the oxidation of
the C-17␤-alcohol to a C-17 ketone must occur via a 17␤ hydrox-
ysteroid dehydrogenase (17␤-HSD). As this only occurs with one
of the four isomers it demonstrates that there is a very pre-
cise stereochemical requirement for the oxidation of the C-17␤
hydroxyl group. This is interesting in that some 17␤-HSD enzymes
show a degree of flexibility in structure with respect to han-
dling steroidal substrates e.g. in bacteria [30] and plant [31], or
an ability to accept steroids in alternative binding conformations
in human cells [32,33]. The 17␤-HSD was unable to oxidise the
three other steroidal diol isomers (11,13,14) all of which underwent
11␤-hydroxylation. It is remarkable that the oxidation of 5␣-
androstan-3␤,17␤-diol (14) did not occur, as it is only structurally
different at C-3 where the alcohol has beta stereochemistry. This
is another example of the importance of single functional groups
[3,6] and their stereochemistry in guiding metabolic fate within this
organism. Interestingly, 5␣-androstan-3␤,17␣-diol (13) demon-
strated the greatest range of monohydroxylated products and if
Transformation of 5␣-androstan-3␣,17␤-diol (15) resulted in
two products. 3␣-Hydroxy-5␣-androstan-17-one (7) was devoid of
the triplet signal (4.05 ppm) generated by the 17␣-H in the start-
ing material indicating that oxidation to a ketone may have taken
place. This was confirmed by inspection of the product 13C NMR
spectrum that contained a resonance signal at 221.10 ppm and is
fully consistent with a 5-ring ketone. Comparison of the 1H and 13
C
NMR spectra of (7) to that of an authentic sample [6] fully supported
the proposed structure (Fig. 3).
3␣-Hydroxy-17a-oxa-D-homo-5␣-androstan-17-one (21) was
identified by comparison of its’ spectra to that of (7) this demon-
strated a significant downfield shift for the 18-methyl protons of

A.C. Hunter et al. / Journal of Steroid Biochemistry & Molecular Biology 122 (2010) 352–358
357
these occurred solely within the 11␤-hydroxylase, it would be con-
sistent with reverse (6␤), inverted normal (7␤) and normal (11␤)
binding. Identical positions of hydroxylation have been observed
following metabolism of the 6-membered ring-D containing lac-
tone 3␤-hydroxy-17a-oxa-D-homo-5␣-androstan-17-one [7]. This
may indicate that during the hydroxylation process the lactone
ring is open, thus generating an alcohol with 17␣-stereochemistry
in remarkably spatial proximity to the 17␣-alcohol of (13). The
metabolite 5␣-androstan-3␤,7␤,17␣-triol was isolated in signifi-
cant yield (55%) presumably demonstrating that optimal binding,
if in the 11␤-hydroxylase, is in the inverted normal binding posi-
tion. The recovered total yields of steroid from the transformation
of compounds (11) and (14) were low compared to those from (13)
and (15) (Table 7). Interestingly the steroids with the low recovery
retained both hydroxyl groups on the same side of the molecule
(3␣,17␣ and 3␤,17␤).
It could be speculated that the unique metabolic handling
resulting in compound (21) may either be directly or indirectly
due to compound (15) having inhibitory activity against the 11␤-
hydroxylase. This notion may be supported from previous work in
which the ring-D lactone in a range of steroidal analogues does not
appear to inhibit 11␤-hydroxylase activity [1,6]. No oxidation of
the 3␣-alcohols to 3-ketones was observed, this is consistent with
previous observations [6] with this organism. Previous studies with
A. tamarii have also demonstrated that 3␣-alcohols and acetates [6]
do not block Baeyer–Villiger oxidation of the C-17 ketone, this also
holds true in the presence of 11␣-hydroxyl groups [3].
[9] D.K. Chakravorty, A.V. Soudackov, S. Hammes-Schiffer, Hybrid quan-
tum/classical molecular dynamics simulations of the proton transfer reactions
catalysed by ketosteroid isomerase: analysis of hydrogen bonding, conforma-
tional motions and electrostatics, Biochemistry 48 (2009) 10608–10619.
[10] U. Dhagat, S. Endo, R. Sumii, A. Hara, O. El-Kabbani, Selectivity determinants of
inhibitor binding to human 20 alpha-hydroxysteroid dehydrogenase: crystal
structure of the enzyme internary complex with coenzyme and potent inhibitor
3,5-dichlorosalicylic acid, J. Med. Chem. 51 (2008) 4844–4848.
[11] N. Mast, S.E. Graham, U. Andersson, I. Bjorkhem, C. Hill, J. Peterson, I.A. Pikuleva,
Cholesterol binding to cytochrome P450 7A1, a key enzyme in bile acid biosyn-
thesis, Biochemistry 44 (2005) 3259–3271.
[12] R Shi, S.X. Lin, Cofactor hydrogen bonding onto the protein main chain is con-
served in the short chain dehydrogenase/reductase family and contributes to
nicotinamide orientation, J. Biol. Chem. 279 (2004) 16778–16785.
[13] D. Mazumder, K. Khan, T.C. Bruice, Computational study of ketosteroid iso-
merase: insights from molecular dynamics simulation of enzyme bound
substrate and intermediate, J. Am. Chem. Soc. 125 (2003) 7553–7561.
[14] F. Henot, R.M. Pollack, Catalytic activity of the D38A mutant of 3-oxo-ꢀ5-
steroid isomerase: recruitment of aspartate-99 as the base, Biochemistry 39
(2000) 3351–3359.
[15] M.A. Massiah, C. Abeygunawardana, A.G. Gittis, A.S. Mildvan, Solution structure
of ꢀ5-3-ketosteroid isomerase complexed with the steroid 19-nortestosterone
hemisuccinate, Biochemistry 37 (1998) 14701–14712.
[16] M. Numazawa, M. Tachibana, Y. Tateda, 4-Oxygenated androst-5-en-17-ones
and their 7-oxo derivatives as aromatase inhibitors, J. Steroid Biochem. Mol.
Biol. 58 (1996) 431–438.
[17] J.R. Hanson, P.B. Hitchcock, A.C. Hunter, The microbiological hydroxylation of
3,16-disubstituted androstanes by Cephalosporium aphidicola, Phytochemistry
49 (1998) 1287–1293.
[18] A.G. Oliveira, P.H. Coelho, F.D. Guedes, G.A.B. Mahecha, R.A. Hess, C.A.
Oliveira, 5␣-Andrsotane-3␤,17␤-diol (3␤-diol) an estrogenic metabolite of
5␣-dihydrotestosterone, is a potent modulator of estrogen receptor ER beta
expression in the ventral prostate of adult rats, Steroids 72 (2007) 914–
922.
[19] H. Pettersson, L. Holmberg, M. Axelson, M. Norlin, CYP7B1-mediated
metabolism of dehydroepiansdrosterone and 5␣-androstane-3␤,17␤-
diol—potential role(s) for estrogen signalling, FEBS J. 275 (2008) 1778–1789.
[20] N. Sugiyama, S. Andersson, R. Lathe, X. Fan, P. Alonso-Magdalena, T. Schwend, I.
Nalvarte, M. Warner, J.A. Gustafsson, Spatiotemporal dynamics of the expres-
sion of estrogen receptors in the postnatal mouse brain, Mol. Psychiatry 14
(2009) 223–232.
[21] M.J. Sikora, K.E. Cordero, J.M. Larious, M.D. Johnson, M.E. Lippman, J.M. Rae,
The androgen metabolite 5␣-androstane-3␤,17␤-diol induces breast cancer
growth via estrogen receptor: implications for aromatase inhibitor resistance,
Breast Cancer Res. Treat. 115 (2009) 289–296.
This study has not only revealed unique effects of stereochem-
istry and polar binding on steroidal diol fate in A. tamarii but it
has also generated a range of novel triols, one in significant yield,
which may be of biological interest, as steroids containing this type
of functionality have distinct activity and/or play important inter-
mediate roles in metabolism [34–37].
[22] Y. Omoto, R. Lathe, M. Warner, J.A. Gustafsson, Early onset of puberty and early
ovarian failure in CYP7B1 knockout mice, Proc. Natl. Acad. Sci. U.S.A. 102 (2005)
2814–2819.
Acknowledgement
[23] S.R. Aspinall, S. Stamp, A. Davison, B.K. Shenton, T.W.J. Lennard, The prolifera-
tive effects of 5-androstene-3␤,17␤-diol and 5␣-dihydrotestosterone on cell
cycle analysis and cell proliferation in MCF7, T47D and MDAMB231 breast
cancer cell lines, J. Steroid Biochem. Mol. Biol. 88 (2004) 37–51.
[24] E.H. Nunlist, I. Dozmorov, Y. Tang, R. Cowan, M. Centola, H.-K. Lin, Partition-
ing of 5␣-androstane-3␣,17␤-diol activated pathways for stimulating human
prostate cancer LNCaP cell proliferation, J. Steroid Biochem. Mol. Biol. 91 (2004)
157–170.
[25] H. Pattersson, J. Lundqvist, E. Oliw, M. Norlin, CYP7B1-mediated metabolism
of 5␣-androstane-3␣,17␤-diol (3␣-Adiol): a novel pathway for potential regu-
lation of the cellular levels of androgens and neurosteroids, Biochim. Biophys.
Acta. Mol. Cell Biol. Lipid 1791 (2009) 1206–1215.
We would like to thank the University of Brighton for financial
support.
References
[1] D.R. Brannon, J. Martin, A.C. Oehlschlager, N.N. Durham, L.H. Zalkow, Transfor-
mation of progesterone and related steroids by Aspergillus tamarii, J. Org. Chem.
30 (1965) 760–762.
[2] A.C. Hunter, N.E. Carragher, Flexibility of the endogenous progesterone lactoni-
sation pathway in Aspergillus tamarii KITA: transformation of a series of cortical
steroid analogues, J. Steroid Biochem. Mol. Biol. 87 (2003) 301–308.
[3] A.C. Hunter, J. Elsom, L. Ross, R. Barrett, Ring-B functionalized androst-4-
en-3-ones and ring-C substituted pregn-4-en-3-ones undergo differential
transformation in Aspergillus tamarii KITA: ring-A transformation with all
C-6 substituted steroids and ring-D transformation with C-11 substituents,
Biochim. Biophys. Acta 1761 (2006) 360–366.
[4] A.C. Hunter, S. Kennedy, S.-J. Clabby, J. Elsom, Fate of novel Quasi reverse
steroidal substrates by Aspergillus tamarii KITA: bypass of lactonisation and an
exclusive role for the minor hydroxylation pathway, Biochim. Biophys. Acta
1734 (2005) 190–197.
[5] A.C. Hunter, E. Coyle, F. Morse, C. Dedi, H.T. Dodd, S.-J. Koussoroplis, Transforma-
tion of 5-ene steroids by the fungus Aspergillus tamarii KITA: mixed molecular
fate in lactonization and hydroxylation pathways with identification of a puta-
tive 3␤-hydroxy-steroid dehydrogenase/ꢀ5-ꢀ4 isomerase pathway, Biochim.
Biophys. Acta 1791 (2009) 110–117.
[6] A.C. Hunter, H. Khuenl-Brady, P. Barrett, H.T. Dodd, C. Dedi, Transformation
of some 3␣-substituted steroids by Aspergillus tamarii KITA reveals stereo-
chemical restriction of steroid binding orientation in the minor hydroxylation
pathway, J. Steroid Biochem. Mol. Biol. 118 (2010) 171–176.
[26] T. Higashi, Y. Nagura, K. Shimada, T. Toyo’oka, Studies on neurosteroids
XXV1. Fluoxetine-evoked changes in rat brain and serum levels of neuroac-
tive androgen, 5␣-androstane-3␣,17␤-diol, Biol. Pharm. Bull. 32 (2009) 1636–
1638.
[27] C.A. Frye, A. Ryan, M. Rhodes, Antiseizure effects of 3␣-androstanediol and/or
17␤-estradiol may involve actions at the estrogen receptor ␤, Epilepsy Behav.
16 (2009) 418–422.
[28] D.N. Kirk, H.C. Toms, C. Douglas, A. White, A survey of the high-field 1H NMR
spectra of steroid hormones, their hydroxylated derivatives and related com-
pounds, J. Chem. Soc. Perkin Trans. 2 (1990) 1567–1594.
[29] J.W. Blunt, J.B. Stothers, 13C NMR of steroids—a survey and commentary, Org.
Magn. Reson. 9 (1977) 439–463.
[30] J. Benach, C. Filling, U.C.T. Oppermann, P. Roversi, G. Bricogne, K.D. Berndt,
H. Jörnvall, R. Ladenstein, Stucture of bacterial 3␤/17␤-hydroxysteroid dehy-
˚
drogenase at 1.2 A resolution:
a model for multiple steroid recognition,
Biochemistry 41 (2002) 14659–14668.
[31] V. Heri, J. Frankenstein, N. Meitinger, F. Muller-Uri, W. Kreis, A(5)-3␤-
hydroxysteroid dehydrogenase (3␤-HSD) from Digitalis lanata. Heterologous
expression and characterisation of the recombinant enzyme, Planta Med. 73
(2007) 704–710.
[32] A. Gangloff, R. Shi, V. Nahoum, S.-X. Lin, Pseudo-symmetry of C-19-steroids,
alternative binding orientations and multispecificity in human estrogenic 17␤-
hydroxysteroid dehydrogenase, FASEB J. 16 (2002), doi:10.1096/fj.02-0026fje.
[33] W. Qui, M. Zhou, F. Labrie, S-X. Lin, Crystal structures of the multispecific 17␤-
hydroxysteroid dehydrogenase type 5: critical androgen regulation in human
peripheral tissues, Mol. Endocrinol. 18 (2004) 1798–1807.
[7] A.C. Hunter, H. Bergin-Simpson, Distinct metabolic handling of 3␤-hydroxy-
17a-oxa-D-homo-5␣-androstan-17-one by the filamentous fungus Aspergillus
tamarii KITA: evidence in support of the steroid/hydroxylase binding hypothe-
sis, Biochim. Biophys. Acta 1771 (2007) 1254–1261.
[8] D.A. Kraut, P.A. Sigala, T.D. Fenn, D. Herschlag, Dissecting the paradoxical effects
of hydrogen bond mutations in the ketosteroid isomerase oxyanion hole, Proc.
Natl. Acad. Sci. U.S.A. 107 (2010) 1960–1965.

358
A.C. Hunter et al. / Journal of Steroid Biochemistry & Molecular Biology 122 (2010) 352–358
[34] A-K.Ö. Farrants, I. Björkhem, J.I. Pedersen, Identification of 3␣,7␣,12␣-
trihydroxy-5␤-cholest-24-enoic acid as an intermediate in the peroxisomal
conversion of 3␣,7␣,12␣-trihydroxy-5␤-cholestenoic acid to cholic acid,
Biochim. Biophys. Acta. Lipids Lipid Metab. 1002 (1989) 198–202.
[35] L.-H. Shan, H.-M. Liu, K.-X. Huang, G.-F. Dai, C. Cao, R.-J. Dong, Synthesis
of 3␤,7␣,11␣-trihydroxy-pregn-21-benzylidene-5-en-20-one derivatives and
their cytotoxic activities, Bioorg. Med. Chem. Lett. 19 (2009) 6637–6639.
[36] A.Y. Reeder, G.E. Joannou, 15␤-Hydroxysteroids (Part IV). Steroids of the human
perinatal period: the synthesis of 3␣,15␤,17␣-trihydroxy-5␣-pregnan-20-one
and it’s A/B-ring configurational isomers, Steroids 60 (1995) 796–801.
[37] S. Christakoudi, D.A. Cowan, N.F. Taylor, A new marker for early diagnosis of
21-hydroxylase deficiency: 3␤,16␣,17␣-trihydroxy-5␣-pregnane-7,20-dione,
J. Steroid Biochem. Mol. Biol. 121 (2010) 574–581.