Metabolic fate of 3α,5-cycloandrostanes in the endogenous lactonization pathway of Aspergillus tamarii KITA

Article abstract of DOI:10.1016/j.phytochem.2015.09.003

A series of 3α,5-cycloandrostane analogues with a range of functionality (6α and 6β alcohols and ketone) at carbon 6 were tested in the endogenous lactonization pathway in Aspergillus tamarii KITA. This metabolic route converts progesterone to testololactone in high yield through a four step enzymatic pathway. To date, no studies have looked at the effect of steroids devoid of polar functionality at carbon 3 and their subsequent metabolic fate by fungi which contain Baeyer-Villiger monooxygenases. Incubation of all of the cycloandrostane analogues resulted in lactonization of ring-D irrespective of C-6 stereochemistry or absence of C-3 functionality. Presence of 6β-hydroxy group and the C-17 ketone was required in order for these analogues to undergo hydroxylation at C-15β position. All metabolites were isolated by column chromatography and were identified by ¹H, ¹³C NMR, DEPT analysis and other spectroscopic data.

Full text of DOI:10.1016/j.phytochem.2015.09.003

Phytochemistry 119 (2015) 19–25
Contents lists available at ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
Metabolic fate of 3
a,5-cycloandrostanes in the endogenous lactonization
pathway of Aspergillus tamarii KITA
A. Christy Hunter a, , Shreyal Patel , Cinzia Dedi , Howard T. Dodd , Richard A. Bryce
⇑
b
b
b
a
a
Manchester Pharmacy School, University of Manchester, Stopford Building, Oxford Road, Manchester M13 9PT, United Kingdom
University of Brighton, School of Pharmacy and Biomolecular Sciences, Huxley Building, Lewes Road, Brighton BN2 4GJ, United Kingdom
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 June 2015
Received in revised form 27 August 2015
Accepted 8 September 2015
Available online 11 September 2015
A series of 3
a
,5-cycloandrostane analogues with a range of functionality (6
a
and 6b alcohols and ketone)
at carbon 6 were tested in the endogenous lactonization pathway in Aspergillus tamarii KITA. This meta-
bolic route converts progesterone to testololactone in high yield through a four step enzymatic pathway.
To date, no studies have looked at the effect of steroids devoid of polar functionality at carbon 3 and their
subsequent metabolic fate by fungi which contain Baeyer–Villiger monooxygenases. Incubation of all of
the cycloandrostane analogues resulted in lactonization of ring-D irrespective of C-6 stereochemistry or
absence of C-3 functionality. Presence of 6b-hydroxy group and the C-17 ketone was required in order for
these analogues to undergo hydroxylation at C-15b position. All metabolites were isolated by column
chromatography and were identified by 1H, C NMR, DEPT analysis and other spectroscopic data.
Ó 2015 Elsevier Ltd. All rights reserved.
Keywords:
Aspergillus tamarii
3a,5-Cyclosteroid
Biocatalysis
Lactonization
Hydroxylation
Biotransformation
13
1
. Introduction
Although a wide range of investigations have tested the flexibil-
ity of the endogenous lactonization pathway within A. tamarii
The fungus Aspergillus tamarii KITA converts progesterone 1 into
testololactone 5 in a high yield four step enzymatic pathway
Brannon et al., 1965). This starts with Baeyer–Villiger oxidation
(Leisch et al., 2011), the metabolism of cycloandrostanes has not
yet been determined. The cyclandrostane analogues used in this
study contain a cyclopropane ring in which all carbon–hydrogen
(
À1
of the C-20 side chain ketone of progesterone 1, generating
testosterone acetate 2 which then undergo hydrolysis to form
testosterone 3. Oxidation of the C-17b-alcohol follows, producing
androst-4-en-3,17-dione 4. Baeyer-Villiger oxidation of the C-17
ketone is the final step this producing testololactone 5. The
pathway reaches a divergent point with testosterone, where 11b-
hydroxytestosterone can be generated from a minor hydroxylation
pathway. To date, the 11b-hydroxylated derivative of testololac-
tone 5 has not been isolated, indicating that the presence of this
functionality may be inhibitory to gaining access to the Baeyer–
Villiger monooxygenase. In contrast to this, the presence of a ring-D
lactone does not inhibit access of these steroids to the 11b-hydrox-
ylase (Hunter and Bergin-Simpson, 2007; Hunter et al., 2010a).
Furthermore, a reversible equilibrium exists, where the C-20
ketone of progesterone 1 is reduced to the 20(R)-alcohol 6 and vice
versa in an equilibrium based reaction. 20(R)-reduction products of
progesterone analogues have been isolated in a number of trans-
formation studies (Hunter and Carragher, 2003; Hunter et al.,
bonds are eclipsed, contributing to strain energy of 117.6 kJ mol
(Liebman and Greenberg, 1976). This results in
a
high
thermodynamic driving force for ring opening reactions, the
presence of which may be used to elucidate the mechanistic basis
of the biocatalytic reactions (Suckling, 1988). Relatively few
studies have determined the metabolic fate of 3a,5-cyclosteroids
in fungi. Those that have (Procházk et al., 1975; Chambers et al.,
1975; Thoa et al., 1978; Bensasson et al., 1999) tended to
observe hydroxylation reactions as opposed to Baeyer–Villiger
oxidation, making the handling of these steroids within the
endogenous lactonization pathway within Aspergillus tamari
(Fig. 1) of particular interest.
In this series of cyclosteroids the effect of C-6 alpha- and beta-
hydroxy as well as keto-functionality will be determined on the
cyclopropane group on ring-A. Previous studies determining
the biocatalysis of steroids by A. tamarii have demonstrated that
the presence of polar functionality at C-6 (Hunter et al., 2006)
drives enzymatic attack to ring-A. These analogues will also
facilitate the determination of the effect of absence of polar
functionality at C-3, as well as reveal the reliance of the endoge-
nous Baeyer-Villiger oxygenase on this group for lactonization of
ring-D (see Fig. 2).
2
006) with this organism.
⇑
Corresponding author.
E-mail address: Christy.hunter@manchester.ac.uk (A.C. Hunter).
http://dx.doi.org/10.1016/j.phytochem.2015.09.003
031-9422/Ó 2015 Elsevier Ltd. All rights reserved.
0

2
0
A.C. Hunter et al. / Phytochemistry 119 (2015) 19–25
O
O
O
OH
O
O
O
(1)
(2)
(3)
(4)
(5)
O
H
OH
Minor 11β-hydroxylation
pathway
(6)
Fig. 1. The endogenous lactonization pathway present in Aspergillus tamari KITA converts progesterone (1) to testosterone acetate (2) via Baeyer–Villiger oxidation, this then
undergoes hydrolysis forming testosterone (3), the 17b-alcohol of which is oxidised to androst-4-en-3,17-dione (4), Baeyer–Villiger oxidation of the C-17 ketone generates
testololactone (5). Progesterone (3) can also undergo reduction via a C-20 keto-alcohol isomerase generating 20(R)-hydroxy-pregna-4-en-3-one (6). A minor hydroxylation
pathway generates 11b-hydroxytestosterone from testosterone (3).
2
. Results
authentic sample. The only product of transformation was isolated
at a concentration of 50% ethyl acetate in light petroleum ether as a
2
2
.1. Products isolated following transformation
gum (203 mg, 14%) and was identified as 17a-oxa-D-homo-3a,5-
1
cycloandrost-6,17-dione 14. The H NMR spectrum of this metabo-
lite contained no new signals however; the 18-methyl signal had
undergone a significant downfield shift (0.43 ppm) which was con-
sistent with insertion of oxygen into ring-D (Hunter et al., 2009,
.1.1. Transformation of 6b-hydroxy-3a,5-cyclo-androstan-17-one 8
Starting material 8 (150 mg, 10%) was recovered from the chro-
matography column at a concentration of 30% ethyl acetate in light
petroleum ether, the structure of which was confirmed by compar-
ison with that of an authentic sample. The first product of transfor-
mation eluted at a concentration of 65% ethyl acetate in petroleum
ether as a gum and was identified as 6b-hydroxy-17a-oxa-D-
1
3
2010b). The C NMR spectrum of the product was also consistent
with this transformation, where the C-13 and C-18 signals had
undergone downfield shifts of 34.79 ppm and 6.36 ppm respec-
tively due to proximity to 17a-oxa functionality. Upfield shifts
associated with ring expansion were observed for C-14
(5.25 ppm), C-16 (7.25 ppm) and C-17 (48.99 ppm) all of which
are consistent with lactone formation on ring-D of the steroid
nucleus (Hunter et al., 2009, 2010b). Accurate mass measurement
fully supported the proposed structure HRMS ESI M+Na+ obsd.
homo-3a,5-cycloandrostane 12 (184 mg, 12%). There were no
1
new signals in this product H NMR but, the 18-methyl signal
had undergone a significant downfield shift (0.43 ppm) consistent
with insertion of oxygen into ring-D (Hunter et al., 2009, 2010b).
1
3
The C NMR spectrum of the product was also consistent with this
transformation where the C-13 ( 36.93 ppm) and C-18 (
.54 ppm) signals had undergone a downfield shifts due to proxim-
D
D
325.178 C19
3
H28NaO req. 325.177.
7
ity to 17a-oxa functionality. Also upfield shifts associated with ring
expansion were observed for C-14 (4.03 ppm), C-16 (6.1 ppm) and
C-17 (47.92 ppm), all of which are consistent with lactone forma-
tion on ring-D of the steroid nucleus (Hunter et al., 2009, 2010b).
Accurate mass measurement fully supported the proposed struc-
2.1.3. Transformation of 6b,17b-dihydroxy-3a,5-cycloandrostane 10
Elution with 70% ethyl acetate in petroleum ether afforded the
starting material (143 mg, 9.5%) which was identified by compar-
ison of its spectral data with that of an authentic compound.
Elution with ethyl acetate afforded the isolation of two products as
ture HRMS ESI M+Na+ obsd. 327.193 C19
H28NaO
req. 327.193
3
gums, namely 6b-hydroxy-17a-oxa-D-homo-3a,5-cycloandrostane
(
see Fig. 3).
12 (198 mg, 13%) and 6b,15b-dihydroxy-3 ,5-cycloandrost-17-one
a
The second product of metabolism eluted at a concentration of
0% ethyl acetate in petroleum ether, this was isolated as a gum
13 (77 mg, 5%). These products are identical to those isolated
following transformation of compound 8. All spectroscopic and
mass spectral data were found to be consistent with authentic
samples (vide supra).
9
and identified as 6b,15b-dihydroxy-3
a,5-cycloandrost-17-one 13
1
(
60 mg, 4%). The H NMR spectrum contained a new resonance sig-
nal at 4.55 ppm (1H, dd) indicating monohydroxylation at an equa-
torial position in the starting material. This was further confirmed
2
.1.4. Transformation of 6
a
,17b-dihydroxy-3
a
,5-cycloandrostane 11
+
by accurate mass measurement HRMS ESI M+Na obsd. 327.193
Transformation of 6
a
,17b-dihydroxy-3
a
,5-cycloandrostane 11
19 3
C H28NaO
req. 327.193. The 13C NMR spectrum of the product
afforded the isolation of one product of metabolism, namely 6
hydroxy-17a-oxa-D-homo-3 ,5-cycloandrostane 15 which, eluted
from the chromatography column at 70% ethyl acetate in petro-
leum ether as a gum (85 mg, 21%). The H NMR spectrum of the
a-
demonstrated b-carbon upfield shifts in the resonance signals for
C-14 (5.72 ppm), C-16 (11.29 ppm) and C-18 (5.16 ppm) and
a
downfield
c
-carbon shifts for C-8 (4.1 ppm) and C-13 (0.28 ppm)
1
C-17 (1.57) which were consistent with literature values (Blunt
and Stothers, 1977) [14] for C-15b hydroxylation.
product was devoid of the 17a-H signal (1H, dd) at 3.89 ppm indi-
cating that the 17b-alcohol was not present, this coupled with the
significant downfield shift (0.4 ppm) for the 18-methyl group indi-
cated that ring-D had undergone lactonization. This was fully sup-
ported following comparison of the starting material 11 and
2.1.2. Transformation of 3
a,5-cycloandrostan-6,17-dione 9
Starting material 9 (367 mg, 24%) eluted from the chromatogra-
phy column at 25% ethyl acetate in light petroleum ether, the
structure of which was confirmed by comparison with that of an
1
3
product 15
C NMR spectra where C-13 (38.11 ppm), C-17
(82.27 ppm) and C-18 (7.58 ppm) resonance signals underwent

A.C. Hunter et al. / Phytochemistry 119 (2015) 19–25
21
O
O
O
(ii)
(i)
(9)
(7)
(8)
HO
OH
O
(iii)
OH
OH
+
(11)
(10)
OH
OH
(
i) Pyridine; CH ClO S; 20hr (ii) CH CH CO; Jones' Reagent;10 min (iii)CH OH; NaBH ; CH CO H; 1hr
3
2
3
3
3
4
3
2
Fig. 2. Synthetic pathway for the synthesis of the 3
cycloandrostan-6,17-dione (9); 6b,17b-dihydroxy-3
a
,5-cycloandrostanes analogues. Analogues are named as follows: 6b-Hydroxy-3
a
,5-cyclo-androstan-17-one (8); 3a,5-
a
,5-cycloandrostane (10); 6 ,17b-dihydroxy-3 ,5-cycloandrostane (11).
a
a
downfield shifts and C-14 (3.97 ppm) and C-16 (1.98 ppm) under-
went upfield shifts. These were consistent with Baeyer–Villiger
oxidation of C-17 ketone as well as difference in resonance values
observed when starting with C-17b alcohol to ring-D lactone
The fungus monohydroxylated the cycloandrostanes exclu-
sively at the C-15b position. This only occurred in the presence
of a 6b-alcohol and 17-keto or 17b-hydroxy functionality in com-
pounds 8 and 10 respectively. It is of note that the C-17b analogue
10 underwent oxidation to form the C-17 ketone, thus indicating
that this functionality was required for entrance into a hydroxyla-
tion pathway. The minor 11b-hydroxylation pathway has always
been observed to act on testosterone 3 (Brannon et al., 1965) which
has a 17b-alcohol. This may suggest that the 15b-hydroxylation is
being performed by a hydroxylase other than the 11b due to the
absence of alcohol or keto functionality at C-3. Alternatively, if
the C-6b alcohol can bind to the 11b-oxidase site it results in a cap-
sized position resulting in the C-15b-proton situated over the C-17
binding position. As mixed function binding/hydroxylation (Jones,
1973; Holland, 1982) functionality has been proposed for hydrox-
ylase enzymes, the C-17 binding region could act as the source of
C-15b-hydroxylation in the 11b-hydroxylase. At present there are
no crystal structures available for steroidal 11b-hydroxylases. In
order to differentiate between the two possibilities, we have used
molecular mechanics to model the 11b-hydroxytestosterone as a
marker for polar functional group positions. Structures of 11b-
(Hunter et al., 2009, 2010b).
3
. Discussion
All of the cycloandrostanes (8, 9, 10, 11) were able to access the
endogenous Baeyer–Villiger monooxygenase present in A. tamarii
KITA that is responsible for generating ring-D steroidal lactones.
Incubation of the diols 10, and 11 confirmed that these
cyclosteroids could enter the lactonization pathway (Fig. 1) and
be selectively handled by the 17b-hydroxy oxidase, generating
the C-17 ketone. The Baeyer–Villiger monooxygenase (and 17b-
oxidase) were not affected by the absence of functionality at C-3 on
the steroid molecule which, has been present in the majority of
studies investigating lactonization of ring-D of steroids in A. tamarii
(Brannon et al., 1965; Hunter et al., 2005; Leisch et al., 2011) and
other fungal species (Al-Aboudi et al., 2008; Kolek et al., 2009;
´
Huang et al., 2010; Zhang et al., 2013; Swizdor et al., 2011, 2013,
hydroxytestosterone and 6b-hydroxy-3a,5-cycloandrostan-17-one
2
2
014; Lobastova et al., 2015) to date. Previously (Hunter et al.,
006), the presence of C-6 functionality on the steroid skeleton
8 were constructed in MOE and then energy minimised. An initial
inspection of the 11b-hydroxy-testosterone (Fig. 4a) and 8 (Fig. 4b)
reveals the structurally striking difference in ring-A conformation
of the two steroids, the former being planar and the latter having
a distorted boat conformation (Cox et al., 1990) (marked in green,
Fig. 4a and b). In order for 8 to bind to the 11b-hydroxylase
it would require the enzyme to have a large hydrophobic
pocket available to accommodate the ring-A of the cyclosteroids.
directed enzymatic attack to ring-A, for the cyclosteroids tested
no structural changes were observed. When attack has occurred
to this centre, it has tended to be reductive, suggesting that the role
of C-6 functionality in the metabolic pathways present in A. tamarii
is solely or partially (3-keto-4-ene functionality present in previous
study) responsible for driving exclusive reductive transformation.

2
2
A.C. Hunter et al. / Phytochemistry 119 (2015) 19–25
O
O
O
O
(i)
+
(12)
(13)
(8)
OH
OH
OH
OH
O
O
O
(i)
(14)
(9)
O
O
OH
O
O
O
(i)
+
(12)
(13)
(10)
OH
OH
OH
OH
OH
O
O
(i)
(15)
(11)
OH
OH
(i) Aspergillus tamarii KITA
Fig. 3. Metabolism of 3
,5-cycloandrostane (12); 6b,15b-dihydroxy-3
cycloandrostane (15).
a
,5-cycloandrostanes following 5 days incubation with Aspergillus tamarii. Isolated metabolites are named as follows: 6b-hydroxy-17a-oxa-D-homo-
3a
a
,5-cycloandrost-17-one (13); 17a-oxa-D-homo-3 ,5-cycloandrost-6,17-dione (14); 6 -hydroxy-17a-oxa-D-homo-3 ,5-
a
a
a
Comparison of the distances between the binding oxygen
functionality on both molecules clearly demonstrates that they
are not compatible (Fig. 4a and b). This adds further weight to
the possibility of a new 15b-steroid hydroxylase enzyme present
in A. tamarii. If the endogenous 11b-steroid hydroxylase can
accommodate ring-A of the cyclosteroid and binding at the
C-3 position is not required, then the 6b-alcohol (if the steroid is
capsized) could bind to the 11b-position (Fig. 4c and d) rendering

A.C. Hunter et al. / Phytochemistry 119 (2015) 19–25
23
access from (11b) and into (1b) the lactonization pathway
Hunter et al., 2010a) with 3 -hydroxy-17-one saturated steroidal
(a)
(b)
(
a
1
7
analogues. Examples of transformation of similar un-saturated
steroids have been reported (Al-Aboudi et al., 2008). For example,
access into an endogenous lactonization pathway has been demon-
4
.99
7
.69
17
11
4
.27
7
.02
strated in Beauveria bassiana, where 11a-hydroxylation is required
´
in both saturated (Swizdor et al., 2014) and un-saturated steroids
1
0.85
15
´
4
.58
(Swizdor et al., 2011) prior to the Baeyer–Villiger monooxygenase
6
transforming the steroids. In contrast to this Fusarium lini can
undergo 11a-hydroxylation of un-saturated steroids following,
but not prior to lactonization (Al-Aboudi et al., 2008).
3
This initial study into the handling of these cycloandrostanes by
A. tamarii has demonstrated the substrates (1–4) can enter both
the lactonization and a hydroxylation pathway, but they require
specific functional group requirements to do so. It is also of note
that cyclosteroids are found as products of endogenous steroid
metabolism in urine (Crawley et al., 2004; Mareck et al., 2008). If
present in nature they may be modified to more polar species
by fungi, resulting in enhanced solubility in water, potentially
increasing their biological activity.
(
c)
(d)
17/15
1
1/6
4
. Experimental
.1. Chemicals and reagents
The 3 ,5-cycloandrostanes were prepared using a previously
4
17/15
11/6
a
reported method (Kovganko and Kashkan, 2001). Comparison of
the reported spectroscopic data as well as accurate mass measure-
Fig. 4. Modelled conformation of (a) 11b-hydroxy-testosterone and (b) 6b-
hydroxy-3 ,5-cyclo-androstan-17-one 8, with inter-oxygen distances indicated in
blue (units in Å) and atom numbering in black. A-ring marked in green. (c) Top and
d) side views of overlay of 11b-hydroxy-testosterone positions 11 and 17, with
capsized 6b-hydroxy-3 ,5-cyclo-androstan-17-one 8 positions 6 and 15 respec-
a
ment confirmed structures 8 (HRMS ESI C19
H
28NaO
2
calc. for M
26NaO calc.
+
(
+Na :311.198 obsd. 311.198) and 9 (HRMS ESI C19
H
2
+
a
for M+Na :309.182 obsd. 309.183). In our hands reduction of 9
with sodium borohydride afforded the diols 10 and 11 as gums
in a ratio of 5:1 which, was in contrast to the mono-keto-alcohols
reported in [13]. DHEA was a kind gift from Professor J. R. Hanson
(University of Sussex), all other chemicals were supplied by the
Aldrich Chemical company (UK); Solvents were of analytical grade;
petroleum ether refers to the fraction with a boiling point of 60–
80 °C. Silica for column chromatography was Merck9385 and TLC
tively. Overlaid oxygen atoms 11/6 and 17/15 are separated by 1.13 and 0.57 Å,
respectively.
the 15b-proton within 1 Å of the C-17 binding site which is well
within the feasible distance (2.5 Å) for radical attack resulting in
hydroxylation (Hirao et al., 2005). This notion is not without merit,
in that an isolated 15b-hydroxylase has been reported capable of
hydroxylating steroids at multiple centres including C-11 (Kiss
et al., 2015).
was performed with Macherey–Nagel AlugramÒ SIL G/UV254
.
Lactonization did not occur in the presence of the 15b-alcohol
suggesting that this hydroxyl group inhibited access to the
Baeyer–Villiger monooxygenase. This would be consistent with
the generation of 11b-hydroxytestosterone which does not
undergo lactonization in A. tamarii (Brannon et al., 1965). In con-
4.2. 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 (Oxoid, UK) slopes (3 days) and maintained
at 4 °C until use. Steroid transformation studies were carried out
in 3% malt extract medium (Oxoid, UK).
trast androst-5-en-3,17-dione has undergone 15a-hydroxylation
followed by ring-D lactonization by Penicillium griseopureum, an
organism in which an 11b-hydroxylation pathway has not been
reported (Huang et al., 2010). Other studies in this fungus have
demonstrated in fully saturated lactones that monohydroxylation
4.3. Conditions of cultivation and transformation
can occur at a wide range of positions (6b, 7b, 11
and Bergin-Simpson, 2007) as well as hydroxylation controlling
a
, 11b) (Hunter
Spores were transferred aseptically in a category 2 biological
safety cabinet into 500 ml Erlenmeyer flasks containing 300 ml of
Table 1
1
3
H NMR data for steroidal starting material and transformation products determined in CDCl .
1
8-H
3
19-H
3
6a-H
Other significant signals
Reference data
6
3
6
6
b-Hydroxy-3
,5-Cycloandrostan-6,17-dione (9)
b,17b-Dihydroxy-3 ,5-cycloandrostane (10)
,17b-Dihydroxy-3 ,5-cycloandrostane (11)
a
,5-cyclo-androstan-17-one (8)
0.93
0.93
1.00
0.93
1.09
1.04
0.73
0.77
3.32 (1H, t, J = 2.8 Hz)
a
a
a
3.21 (1H, t, J = 2.8 Hz) 3.58 (1H, t, J = 8.7 Hz, 17 -H)
a
a
3.66 (1H, t, J = 8.6 Hz, 17a-H); 3.91 (1H, dd, J = 11.4 Hz, J = 4.4 Hz, 6b-H)
Transformation products
6
6
1
6
b-Hydroxy-17a-oxa-D-homo-3
b,15b-Dihydroxy-3
7a-oxa-D-homo-3
-Hydroxy-17a-oxa-D-homo-3
a
,5cycloandrostane (12) 1.36
1.04
1.11
0.99
0.90
3.33 (1H, t, J = 2.9 Hz)
3.37 (1H, t, J = 2.8 Hz) 4.55 (1H, dd, J = 6 Hz, J = 4 Hz, 15a-H)
a
a
,5-cycloandrost-17-one (13)
,5-cycloandrost-6,17-dione (14)
1.23
1.36
a
a
,5-cycloandrostane (15) 1.33
3.89 (1H, dd, J = 4.37 Hz, J = 1.40 Hz, 6b-H)

2
4
A.C. Hunter et al. / Phytochemistry 119 (2015) 19–25
Table 2
1
3C NMR data for starting material and transformation products determined in CDCl
3
.
Carbon atom
Compounds
(
8)
(9)
33.47
(10)
(11)
(12)
33.22
(13)
33.17
(14)
33.46
(15)
32.82
24.95
20.08
6.74
1
2
3
4
5
6
7
8
9
32.00
23.57
22.96
10.50
37.67
72.04
34.76
28.34
46.70
41.82
20.56
30.57
46.70
50.11
20.75
34.76
32.18
23.96
21.26
10.64
41.94
72.58
37.81
29.90
46.79
42.20
22.35
35.91
46.79
49.96
23.20
29.50
80.91
10.26
20.26
32.76
25.03
18.69
6.66
25.84
35.45
11.80
46.36
208.32
43.49
34.34
46.28
46.70
22.14
31.41
47.83
51.97
21.67
35.73
219.90
13.76
19.67
24.92
24.23
11.66
38.49
73.32
35.62
32.69
46.65
42.97
23.51
39.48
83.63
46.08
19.78
28.66
171.85
20.25
20.13
24.88
25.75
11.79
37.07
66.80
35.33
24.24
48.43
41.57
21.74
33.17
46.42
55.83
73.58
46.05
218.20
17.87
20.15
25.76
35.85
12.09
46.03
207.59
43.20
36.99
44.95
45.91
23.43
39.04
82.62
46.72
19.70
28.48
170.91
20.12
19.36
39.84
67.06
39.68
34.98
47.56
43.14
23.41
36.83
45.01
50.76
22.74
30.50
81.85
11.17
18.00
39.46
66.76
38.65
37.58
46.04
44.78
23.83
39.54
83.12
46.79
19.96
28.59
171.12
18.75
17.68
10
11
12
13
14
15
16
17
18
19
219.77
12.71
18.98
Table 3
À1
Significant metabolite absorption signals cm
.
Compound
AOH
>C@O
Lactone
6
6
1
6
b-Hydroxy-17a-oxa-D-homo-3
b,15b-Dihydroxy-3 ,5-cycloandrost-17-one (13)
7a-Oxa-D-homo-3 ,5-cycloandrost-6,17-dione (14)
-Hydroxy-17a-oxa-D-homo-3 ,5-cycloandrostane (15)
a
,5cycloandrostane (12)
3451 br
3374 br
1709
1729 br
1722
a
a
1677
a
a
3353 br
1728 br
sterile media and were incubated for 72 h at 25 °C. The cultures
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 fungus 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
further 5 days after which the metabolites were extracted from
the broth.
(12 ml) was added. The solvent was removed in vacuo and the
resultant gum was adsorbed to silica and chromatographed on a
column using light petroleum ether (60–80) and ethyl acetate as
solvents. This afforded 6b,17b-dihydroxy-3
(10) (2 g) and ,17b-dihydroxy-3 ,5-cycloandrostane (11)
(400 mg) as gums, both of which afforded identical accurate mass
a,5-cycloandrostane
6
a
a
+
measurements (HRMS ESI C19
obsd. 311.198).
2
H28NaO calc. for M+Na :311.198
4.6. Analysis and identification of metabolites
4.4. Extraction and identification of metabolites
Characteristic splitting patterns (Blunt and Stothers, 1977) and
The fungal mycelium was separated from the broth by filtration
1
13
shift values (Kirk et al., 1990) in the H and C NMR spectra in
comparison to the starting compounds were used to determine
metabolite structure in combination with DEPT analysis to identify
the nature of the carbon (Tables 1 and 2). Spectra were recorded on
a Bruker WM 360 Spectrometer, all samples were analysed in deu-
teriochloroform using tetramethylsilane as the internal standard.
High resolution mass spectrometry (HRMS) was determined in
electrospray ionization (ESI) mode using a Bruker Daltonics micrO-
TOF spectrometer. Infra-red absorption spectra (Table 3) were
recorded directly on a Nicolet avatar 320 FT-IR fitted with a Smart
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 twice with
ethyl acetate (1.5 L). The organic extract was dried over sodium
sulphate 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
concentrations of ethyl acetate in petroleum ether. The solvent
was collected in aliquots (10 ml) and analysed by thin layer chro-
matography (TLC) to identify the separated metabolite fractions.
The solvent systems used for running the TLC plates were 50:50
petroleum ether in ethyl acetate or pure ethyl acetate. A 50:50 sul-
phuric acid in methanol spray was used to develop the TLC plates.
Ò
Golden Gate .
4.7. Molecular modelling
4.5. Reduction of 3
a,5-cycloandrostan-6,17-dione 9
The MOE modelling package (Chemical Computing Group Inc.,
1
010 Sherbooke St. West, Suite #910, Montreal, QC, Canada, H3A
3
a
,5-Cycloandrostan-6,17-dione 9 (2.4 g) dissolved in methanol
64 ml) was treated with NaBH (0.28 g). This mixture was stirred
and kept at 0 °C for 1 h in an ice bath, following which acetic acid
2R7, 2015.) was used to construct steroid molecules and
subsequently energy minimise their structures using the MMFF94x
force field (Halgren, 1996).
(
4

A.C. Hunter et al. / Phytochemistry 119 (2015) 19–25
25
of a putative 3b-hydroxy-steroid dehydrogenase/D5
Biochim. Biophys. Acta 1791, 110–117.
Hunter, A.C., Khuenl-Brady, H., Barrett, P., Dodd, H.T., Dedi, C., 2010a.
Transformation of some 3 -substituted steroids by Aspergillus tamarii KITA
-
D4 isomerase pathway.
References
Al-Aboudi, A., Mohammad, M.Y., Musharraf, S.G., Choudhary, M.I., Atta-ur-Rahman,
a
2
008. Microbial transformation of testosterone by Rhizopus stolonifer and
reveals stereochemical restriction of steroid binding orientation in the minor
hydroxylation pathway. J. Steroid Biochem. Mol. Biol. 118, 171–176.
Hunter, A.C., Collins, C., Dodd, H.T., Dedi, C., Koussoroplis, S.-J., 2010b.
Transformation of a series of saturated isomeric steroidal diols by Aspergillus
tamari KITA reveals a precise stereochemical requirement for entrance into the
lactonization pathway. J. Steroid Biochem. Mol. Biol. 122, 352–358.
Jones, E.R.H., 1973. The microbiological hydroxylation of steroids and related
compounds. Pure Appl. Chem. 33, 39–52.
Fusarium lini. Nat. Prod. Res. 22, 1498–1509.
Bensasson, C.S., Hanson, J.R., Le Huerou, Y., 1999. The microbiological hydroxylation
of 3a,5-cycloandrostanes by Cephalosporium aphidicola. Phytochemistry 52,
1
279–1282.
Blunt, J.W., Stothers, J.B., 1977. 13C NMR of steroids-a survey and commentary. Org.
Magn. Reson. 9, 439–463.
Brannon, D.R., Martin, J., Oehischlager, A.C., Durham, N.N., Zalkow, L.H., 1965.
Transformation of progesterone and related steroids by Aspergillus tamarii. J.
Org. Chem. 30, 760–762.
Chambers, V.E.M., Denny, W.A., Jones, E.R.H., Meakins, J.G.D., Miners, J.O., Pinhey, J.
T., Wilkins, A.L., 1975. Microbiological hydroxylation. Part XVII.
Kirk, D.N., Toms, H.C., Douglas, C., White, K.A., 1990. A survey of the high-field 1
H
NMR spectra of steroid hormones, their hydroxylated derivatives and related
compounds. J. Chem. Soc. [Perkin 2] 2, 1567–1594.
Kiss, F.M., Lundemo, M.T., Zapp, J., Woodley, J.M., Bernhardt, R., 2015. Process
development for the production of 15b-hydroxycyproterone acetate using
Bacillus megaterium expressing CYP106A2 as whole-cell biocatalyst. Microb. Cell
Fact. 14, 28. http://dx.doi.org/10.1186/s12934-015-0210-z.
Kolek, T., Szpineter, A., S´ wizdor, A., 2009. Studies on Baeyer–Villiger oxidation of
steroids: DHEA and pregnenolone D-lactonization pathways in Penicillium
camemberti AM83. Steroids 74, 859–862.
Kovganko, N.V., Kashkan, Z.N., 2001. Reactivity of hydroxyl keto groups on C-6 and
C-Hydroxylation of 17-oxo-5
androstanes by the fungus Calonectria decora. J. Chem. Soc. [Perkin 1] 1,
359–1363.
a-androstanes and 17-oxo-3a,5-cyclo-5a-
1
Cox, P.J., MacManus, S.M., Gibb, B.C., Nowell, I.W., Turner, A.B., Howie, R.A., 1990.
Crystal and molecular structure of 3,5-cycloandrostan-6b-ol-17-one. J.
Crystallogr. Spectrosc. Res. 20, 415–418.
Crawley, A.T., Hine, E.R., Trout, G.J., George, A.V., Kazlauskas, R., 2004. Searching for
new markers of endogenous steroid administration in athletes: ‘‘looking
outside the metal box”. Forensic Sci. Int. 143, 103–114.
C-17 of 3a,5a-cycloandrostanes. Chem. Nat. Comp. 37, 47–51.
Leisch, H., Morley, K., Lau, P.C.K., 2011. Baeyer–Villiger monooxygenases: more than
just green chemistry. Chem. Rev. 111, 4165–4222.
Halgren, T.A., 1996. Merck molecular force field. I. Basis, form, scope,
parameterization, and performance of MMFF94. J. Comput. Chem. 17,
Liebman, J.F., Greenberg, A., 1976. Survey of strained organic-molecules. Chem. Rev.
4
90–519.
76, 311–365.
Hirao, H., Kumar, D., Thiel, W., Shaik, S., 2005. Two states and two more in the
mechanisms of hydroxylation and epoxidation by cytochrome P450. J. Am.
Chem. Soc. 127, 13007–13018.
Holland, H.L., 1982. The mechanism of microbiological hydroxylation of steroids.
Chem. Soc. Rev. 11, 371–395.
Huang, L.-H., Li, J., Xu, G., Zhang, X.-H., Wang, Y.-G., Yin, Y.-L., Liu, H.-M., 2010.
Biotransformation of dehydroepiandrosterone (DHEA) with Penicillium
griseopurpureum Smith and Penicillium glabrum (Wehmer) Westling. Steroids
Lobastova, T.G., Khomutov, S.M., Donova, M.V., 2015. Formation of hydroxylated
steroid lactones from dehydroepiandrosterone by Spicaria fumoso-rosea F-881.
Appl. Biochem. Microbiol. 51, 180–187.
Mareck, U., Geyer, H., Opfermann, G., Thevis, M., Schanzer, W., 2008. Factors
influencing the steroid profile in doping control analysis. J. Mass Spectrom. 43,
877–891.
Procházk, Zˇ ., Budĕšinsk y´ , M., Prekajsk, P., 1975. Microbial oxygenation of 6b-
hydroxy-3 ,5-cyclo-5 -androstan-17-one with Rhizopus nigricans. Collect.
a
a
7
5, 1039–1046.
Czech. Chem. Commun. 39, 982–992.
Hunter, A.C., Carragher, N.E., 2003. Flexibility of the endogenous progesterone
lactonization pathway in Aspergillus tamarii KITA: transformation of a series of
cortical steroid analogues. J. Steroid Biochem. Mol. Biol. 87, 301–308.
Hunter, A.C., Kennedy, S., Clabby, S.-J., Elsom, J., 2005. 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
Suckling, C.J., 1988. The cyclopropyl group in studies of enzyme mechanism and
inhibition. Angew. Chem. Int. Ed. Engl. 27, 537–552.
S´ wizdor, A., Kolek, T., Panek, A., Bialoniska, A., 2011. Microbial Baeyer–Villiger
oxidation of steroidal ketones using Beauveria bassiana: presence of an 11a-
hydroxyl group essential to generation of D-homo lactones. Biochim. Biophys.
Acta Mol. Cell Biol. Lipids 1811, 253–262.
1
734, 190–197.
Hunter, A.C., Elsom, J., Ross, L., Barrett, R., 2006. Ring-B functionalized androst-4-en-
-ones and ring-C substituted pregn-4-en-3-ones undergo differential
S´ wizdor, A., 2013. Baeyer–Villiger oxidation of some C19 steroids by Penicillium
lanosocoeruleum. Molecules 18, 13812–13822.
S´ wizdor, A., Panek, A., Milecka-Tronina, N., 2014. Microbial Baeyer–Villiger
3
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, 360–366.
oxidation of
5a
-steroids using Beauveria bassiana.
A
stereochemical
requirement for the 11
Steroids 82, 44–52.
a-hydroxylation and the lactonization pathway.
Hunter, A.C., Bergin-Simpson, H., 2007. Distinct metabolic handling of 3b-hydroxy-
Thoa, H.K., Procházk, Zˇ ., Budĕšinsk y´ , M., Ko cˇ ovsk y´ , P., 1978. Hydroxylation of 6b-
1
7a-oxa-D-homo-5a-androstan-17-one by the filamentous fungus Aspergillus
methoxy-3a,5-cyclo-5a-androstan-17-one with Rhizopus nigricans. Collect.
Czech. Chem. Commun. 43, 2305–2311.
Zhang, H., Ren, J., Wang, Y., Sheng, C., Wu, Q., Diao, A., Zhu, D., 2013. Effective multi-
step functional biotransformations of steroids by a newly isolated Fusarium
oxysporum SC1301. Tetrahedron 69, 184–189.
tamarii KITA: evidence in support of steroid/hydroxylase binding hypothesis.
Biochim. Biophys. Acta 1771, 1254–1261.
Hunter, A.C., Coyle, E., Morse, F., Dedi, C., Dodd, H.T., Koussoroplis, S.-J., 2009.
Transformation of 5-ene steroids by the fungus Aspergillus tamarii KITA: mixed
molecular fate in lactonization and hydroxylation pathways with identification