The generation of a steroid library using filamentous fungi immobilized in calcium alginate Dedicated to the memory of Professor Sir John W. Cornforth, University of Sussex (1917-2013).

Article abstract of DOI:10.1016/j.molcatb.2015.11.026

Four fungi, namely, Rhizopus oryzae ATCC 11145, Mucor plumbeus ATCC 4740, Cunninghamella echinulata var. elegans ATCC 8688a, and Whetzelinia sclerotiorum ATCC 18687, were subjected to entrapment in calcium alginate, and the beads derived were used in the biotransformation of the steroids 3β,17β-dihydroxyandrost-5-ene (1) and 17β-hydroxyandrost-4-en-3-one (2). Incubations performed utilized beads from two different encapsulated fungi to explore their potential for the production of metabolites other than those derived from the individual fungi. The investigation showed that steroids from both single and crossover transformations were typically produced, some of which were hitherto unreported. The results indicated that this general technique can be exploited for the production of small libraries of compounds.

Full text of DOI:10.1016/j.molcatb.2015.11.026

Journal of Molecular Catalysis B: Enzymatic 125 (2016) 16–24
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
The generation of a steroid library using filamentous fungi
immobilized in calcium alginate
Patrice C. Pearta, William F. Reynoldsb, Paul B. Reesea,∗
a Department of Chemistry, University of the West Indies, Mona, Kingston 7, Jamaica
b Department of Chemistry, University of Toronto, Toronto, Ontario M5S 3H6, Canada
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 4 July 2015
Received in revised form
29 November 2015
Accepted 30 November 2015
Available online 6 January 2016
Four fungi, namely, Rhizopus oryzae ATCC 11145, Mucor plumbeus ATCC 4740, Cunninghamella echinulata
var. elegans ATCC 8688a, and Whetzelinia sclerotiorum ATCC 18687, were subjected to entrapment in
calcium alginate, and the beads derived were used in the biotransformation of the steroids 3␤,17␤-
dihydroxyandrost-5-ene (1) and 17␤-hydroxyandrost-4-en-3-one (2). Incubations performed utilized
beads from two different encapsulated fungi to explore their potential for the production of metabolites
other than those derived from the individual fungi. The investigation showed that steroids from both
single and crossover transformations were typically produced, some of which were hitherto unreported.
The results indicated that this general technique can be exploited for the production of small libraries of
compounds.
Dedicated to the memory of Professor Sir
John W. Cornforth, University of Sussex
(1917–2013).
© 2016 Elsevier B.V. All rights reserved.
Keywords:
Steroid
Hydroxylation
Biotransformation
Immobilization
Chemical library
1. Introduction
be symbiotic. For example, one might be oxygen consuming while
the other is oxygen producing, as in the case of Cephalosporium acre-
The concept of using a number of microorganisms in a single
fermentation vessel for the bioconversion of a single substrate to
multiple products is not new. Unfortunately, such an environment
would lead to inhibitory competition, where either one microor-
ganism prevails whilst the other dies, or the growth of both suffers.
Presumably even if the fermentation were successful, the con-
ditions would not be highly reproducible. Additionally, as each
microorganism has different nutritional requirements, it could be
difficult to find a suitable medium to perform the fermentation [1].
Employing immobilized mycelial cells could avert these problems.
With each microorganism trapped in its own sphere, the possibility
of negative competition would be reduced or eliminated.
monium (an oxygen consuming fungus) and Chlorella pyrenoidosa
(an oxygen producing alga). Both were co-immobilized and used
for the production of cephalosporin C [2]. A similar mixed culture
system was previously used for the production of ␤-lactam antibi-
otics [3]. Another study, with entrapped cells of Chlorella vulgaris
and Provedencia sp., as well as Aspergillus nidulans and Proveden-
cia sp., were used produce ␣-keto acids from amino acids [4]. A
mixed culture-entrapped preparation of Chlorella pyrenoidosa and
Glucanobacter oxydans was employed in the bioconversion of glyc-
erol to dihydroxyacetone, in an effort to study the effects that
co-immobilization with oxygen-producing algae had on the trans-
formation yields [5].
Most of the applications of cell immobilization have been
concentrated on monoculture bioconversions. It is, therefore, of
immense interest to explore the potential of mixed cultures. There
have been previous reports of co-immobilized biocatalysts [1]. Typ-
ically, the microorganisms employed for such an objective tend to
It has been shown that immobilized cell bioconversions par-
allel those of the free cell fermentations to some extent [6,7]. A
progression from this would involve the mixing of beads of immo-
bilized fungus A in water with those of mould B in a single vessel,
followed by feeding of the appropriate substrate. The expectation
was the isolation of additional metabolites that were different from
those obtained from the individual single immobilized cell fermen-
tations. Each of these new metabolites would be the product of
“crossover”; that is, a compound formed when the transformed
∗
Corresponding author. Fax: +876 977 1835.
E-mail address: paul.reese@uwimona.edu.jm (P.B. Reese).
http://dx.doi.org/10.1016/j.molcatb.2015.11.026
1381-1177/© 2016 Elsevier B.V. All rights reserved.

P.C. Peart et al. / Journal of Molecular Catalysis B: Enzymatic 125 (2016) 16–24
17
Table 1
13C NMR data (␦) for compounds 2a–9a determined in CDCl3.
Table 2
13C NMR data (␦) for compounds 10a–16a and 6 determined in CDCl3.
Carbon 2a
3a
4a
5a
6a
7a
8a
9a
Carbon 10a
11a
12a
13a
14a
15a
16a
6
1
2
3
4
5
6
7
8
36.1
35.8
199.9 73.0
124.4 37.8
36.5
27.5
36.6
27.5
73.1
37.5
36.5
27.4
72.9
37.8
35.8
33.9
35.5
33.9
36.5
27.7
35.5
33.9
198.8
126.5
1
2
3
4
5
6
7
8
35.9
27.4
72.2
38.0
166.0 76.7
126.7 72.7
201.8 71.4
48.8
44.4
38.6
20.0
27.9
46.9
80.4
35.2
27.2
80.5
16.1
17.2
21.2
21.4
26.0
24.4
70.6
31.2
31.8
26.6
69.9
35.9
75.8
73.0
71.6
34.2
39.4
39.1
20.4
36.3
42.8
44.3
23.1
27.3
82.2
11.9
16.5
20.9
21.0
21.1
33.0
27.0
70.4
37.0
75.7
76.0
31.3
30.2
48.0
40.3
70.3
43.5
42.9
48.9
24.4
27.4
81.6
12.8
16.4
21.2
21.4
21.5
21.9
31.9
26.6
70.1
37.2
75.9
74.9
73.0
35.6
44.2
38.2
20.8
36.8
43.5
49.1
25.3
27.7
82.4
12.3
17.2
21.2
21.3
21.4
21.5
31.9
26.6
70.3
37.0
75.9
76.0
30.1
26.7
45.1
38.9
20.4
38.0
42.7
53.0
71.7
38.3
81.2
14.1
16.4
20.8
21.1
21.3
21.3
32.0
26.6
70.4
36.9
74.9
76.2
25.8
34.0
38.2
38.9
19.5
29.4
47.1
83.0
32.3
26.9
81.0
16.1
16.3
21.1
21.3
21.4
170.0
35.7
33.9
199.4 198.7 72.9
124.1 125.6 37.5
170.2 165.5 144.4 166.3
201.0
123.6
172.8
32.8
26.2
38.7
46.8
39.0
19.0
28.6
47.0
83.3
32.3
29.2
78.3
15.0
17.3
–
171.4 146.7 144.2 146.9
33.2
31.9
34.3
54.1
39.0
20.9
37.0
42.9
50.7
23.9
27.9
82.9
12.4
17.8
21.6
120.6 122.0 120.4
32.5
26.2
38.6
46.7
38.7
19.6
28.9
46.8
82.9
32.8
26.9
80.9
15.9
17.3
21.3
38.1
75.5
39.5
50.5
37.9
20.5
36.3
43.1
49.2
25.3
27.6
81.9
12.0
17.2
21.2
121.9 37.3
67.6
35.6
43.1
37.3
20.2
36.1
42.2
43.7
23.3
27.4
82.5
11.5
18.2
21.2
75.3
36.3
47.9
36.3
20.5
36.5
42.7
49.8
24.5
27.5
82.2
11.8
18.9
21.1
21.3
21.5
67.1
35.5
43.3
37.5
19.9
30.9
47.3
44.7
21.8
35.7
220.3
13.1
18.2
21.3
21.4
170.4
74.9
36.2
48.1
36.7
20.4
31.1
47.7
50.6
23.1
35.7
70.3
38.2
46.4
38.4
20.4
36.1
42.5
44.8
22.9
27.3
36.6
37.0
41.0
20.9
36.1
42.6
44.7
23.1
27.3
82.3
11.7
16.9
21.1
21.1
9
9
10
11
12
13
14
15
16
17
18
19
OAc
OAc
OAc
OAc
OAc
OAc
10
11
12
13
14
15
16
17
18
19
OAc
OAc
OAc
OAc
OAc
OAc
OAc
OAc
220.3 82.2
13.6
19.0
21.4
21.6
11.8
17.1
21.1
21.2
171.6 21.3
171.2 21.6
–
–
–
–
–
–
–
–
–
–
–
21.4
–
–
–
–
170.2 170.3 170.5
171.1 170.9 171.2
–
–
170.2 21.4
171.2 169.3 21.4
–
–
–
–
170.4 170.3 170.6
170.7 170.9
171.3 171.1
–
–
–
–
–
–
170.9 168.7 170.1
171.4 168.8 170.5
–
–
170.2 169.7 170.7
170.3 170.3 171.4
170.8 170.7
171.2 171.0
170.5 170.6
171.4 171.2
–
–
congener from one fungus would be used as the substrate by the
second microorganism. This would lead to the production of multi-
ple compounds. The results of our experiments show this to be the
case. In previous work [6,7]. we reported on the use of six encap-
sulated fungi (Rhizopus oryzae, Mucor plumbeus, Cunninghamella
echinulata var. elegans, Aspergillus niger, Phanerochaete chrysospo-
rium and Whetzelinia sclerotiorum in the transformation of two
steroidal substrates, 3␤,17␤-dihydroxyandrost-5-ene (1) and 17␤-
hydroxyandrost-4-en-3-one (2). In general, it was observed that the
transformations of 1 and 2 by encapsulated A. niger and P. chrysospo-
rium gave modest results. Therefore, only R. oryzae, M. plumbeus,
C. echinulata var. elegans and W. sclerotiorum were chosen for the
crossover experiments. 3␤,17␤-Dihydroxyandrost-5-ene (1) and
17␤-hydroxyandrost-4-en-3-one (2) were used as substrates, as
the current work is an extension of earlier studies [6,7]. With four
fungi available for crossover, there were six possible permutations
with each substrate, giving a total of 12 possible mixed bead fer-
mentations. Once it was determined that one could generate new
compounds from 1 using four combinations of fungi, only two
mixed bead experiments using 2 as substrate were attempted.
Products of transformation derived from 1 were acetylated. This
aided in their purification. It also generated compounds that were
readily soluble in CDCl3. Furthermore, comparison of NMR spec-
tra (for structure elucidation purposes) was facilitated when the
biotransformation products from 1 had been derivitized in this way.
Table 3
13C NMR data (␦) for compounds 7, 9 and 17–22 determined in CDCl3.
Carbon
7
9
17
18
19
20
21
22
23
1
2
3
4
5
6
7
8
35.7
33.9
35.5
34.0
37.1
34.2
37.4
34.2
37.2
34.3
37.4
34.2
33.9
33.8
35.6
34.0
35.4
33.9
199.3 200.7 200.3 199.9 200.4
124.6 126.3 126.5 124.8 126.4
167.5 170.7 167.9 170.1 168.3
202.1 199.7 199.2 198.5
124.2 123.7 126.4 127.2
173.9 163.7 166.7 166.4
42.3
74.9
43.1
50.7
38.0
20.6
36.3
43.5
49.9
26.3
30.7
81.1
11.1
17.3
–
41.2
67.5
40.0
45.4
38.8
20.6
36.2
42.9
45.1
22.8
29.9
81.3
11.0
17.2
–
72.8
37.2
29.6
53.7
38.1
20.3
31.3
47.7
50.9
21.7
35.8
33.4
30.3
34.6
59.2
40.0
68.7
43.0
48.0
50.1
21.7
35.7
73.0
38.1
29.8
53.8
38.1
20.6
36.4
42.9
50.6
23.3
30.5
34.0
31.4
35.6
59.1
40.4
69.5
48.0
43.6
50.0
23.4
29.9
80.8
12.3
18.4
–
128.0 37.4
140.4 70.5
41.0
67.1
39.3
45.4
38.5
20.1
30.9
47.3
45.6
21.3
35.6
220.2
13.5
17.0
–
37.6
50.7
36.1
20.3
36.3
43.8
48.2
23.0
30.4
81.3
11.0
16.3
–
38.6
45.5
38.4
20.5
36.0
42.9
45.0
22.7
30.4
81.4
10.9
17.2
21.2
170.6
9
10
11
12
13
14
15
16
17
18
19
OAc
OAc
220.5 218.4 81.7
13.8
19.6
–
14.7
18.4
–
11.1
19.6
–
–
–
–
–
–
–
–
–
data for compounds that were not previously fully character-
ized are in Tables S8–S11 (Supplementary data). Optical rotations
were performed using a PerkinElmer 241 MC polarimeter and
solutions were prepared using CH2Cl2 unless otherwise stated.
Purifications were done by column chromatography using sil-
ica gel (230–400 mesh) as the stationary phase. Additionally,
preparative thin layer chromatography (PLC) glass backed plates
with silica gel (60 Å, 250 ␮m and 1000 ␮m thicknesses) were
used. Thin layer chromatographic (TLC) analyses were done using
polyester backed plates. Both the TLC and PLC plates were visu-
alized under ultraviolet light or by spraying with ammonium
molybdate-sulfuric acid or methanol-sulfuric acid reagents. After
spraying the plates were warmed for colour development using a
heat gun. Petrol refers to the petroleum fraction boiling between
60 and 80◦. 3␤-Hydroxyandrost-5-en-17-one (dehydroepiandros-
terone, DHEA) was obtained from Productos Químicos Naturales,
2. Experimental
2.1. General procedure
Melting points were obtained using a Reichert Hot Stage
melting point apparatus and are uncorrected. Infrared data was
acquired using a PerkinElmer Fourier transform infrared spec-
trophotometer 1000 using sodium chloride disks. NMR spectra
were obtained on Bruker Avance 200, Bruker Avance 500 and Var-
ian Unity 500 spectrometers. Compounds were analyzed using
CDCl3, unless otherwise stated, with tetramethylsilane as an inter-
nal standard. 13C NMR data for all compounds are reported in
Tables 1–3 . 1D and 2D NMR data for new compounds are
reported in Tables S1–S7 (Supplementary data). 1D and 2D NMR

18
P.C. Peart et al. / Journal of Molecular Catalysis B: Enzymatic 125 (2016) 16–24
S.A. de C.V. (Orizaba, Mexico). 3␤,17␤-Dihydroxyandrost-5-ene (1)
was prepared from DHEA in methanol using sodium borohydride.
17␤-Hydroxyandrost-4-en-3-one (testosterone) (2) was acquired
from Steraloids, Inc. (Wilton, NH, USA). The fungi used for these
experiments were obtained from the American Type Culture Col-
lection, Rockville, MD, USA.
2.3. Preparation of immobilized fungal cells
One slant was used to inoculate four Erlenmeyer flasks each con-
taining 125 mL liquid culture medium. The fungus was allowed to
grow for 3 d. At the end of the incubation period the cells were
harvested by filtration. The cells were suspended in water (40 mL)
and 3% sodium alginate solution (140 mL) was added. The cells
were macerated for 1 min at 8000 rpm using a IKA Ultra-Turrax T25
homogenizer. The cell-alginate suspension was then added drop-
wise to a stirred chilled solution of 0.1 M calcium chloride (200 mL).
The alginate beads formed were allowed to harden for 30 min in
the aqueous calcium chloride solution. The calcium chloride was
decanted and the beads were rinsed with water. The beads were
stored under water at 4◦ until used.
2.2. Media for growth of fungi
The media and culture conditions for the following fungi have
already been reported [6,7]: Rhizopus oryzae ATCC 11145 [8], M.
plumbeus ATCC 4740 [9], C. echinulata var. elegans ATCC 8688a [10]
and W. sclerotiorum ATCC 18687 [11].

P.C. Peart et al. / Journal of Molecular Catalysis B: Enzymatic 125 (2016) 16–24
19
2.4. Mixed culture immobilized cell fermentation conditions
2.5. Biotransformations using 3ˇ,17ˇ-dihydroxyandrost-5-ene
(1)
Equal portions of alginate beads of two fungi immobilized
using the method above were placed into four 500 mL Erlenmeyer
flasks, each containing water (125 mL). The substrate (200 mg) in
ethanol (5 mL) was distributed among the flasks. The flasks were
shaken at 180 rpm for 5 d. After the fermentation was complete
the aqueous solution was decanted from the beads and the former
was extracted using ethyl acetate (2 × 300 mL). The organic solu-
tion was dried using sodium sulfate, filtered, and the solvent was
removed in vacuo. The residue was analyzed by TLC and purified
by column chromatography. The products from the transforma-
tions of 3␤,17␤-dihydroxyandrost-5-ene (1) were poorly resolved.
Acetylation of these metabolites aided in their purification by chro-
matography.
2.5.1. Products from Rhizopus oryzae and Cunninghamella
echinulata var. elegans
Partial purification of the extract (177 mg) by column chro-
matography using acetone/dichloromethane (1:19 v/v) yielded
the fed substrate (1) (58.6 mg). Further purifications using the
same solvent system, followed by acetylation, afforded four
metabolites. The first was 3␤,7␣,17␤-trihydroxyandrost-5-ene
(3). This was characterized as the triacetate (3a) (43.6 mg),
which crystallized from acetone-methanol as plates, Rf = 0.89, ace-
tone/dichloromethane (1:19 v/v), m.p. 142−144◦ [␣]D −176.8◦
(c 0.22), lit [12]. 156–158◦, [␣]D −152◦; IR: ꢀmax 1747, 1723,
1246 cm−1 1H NMR: ␦ 0.80 (3H, s, H-18), 1.03 (3H, s, H-19), 2.04
;

20
P.C. Peart et al. / Journal of Molecular Catalysis B: Enzymatic 125 (2016) 16–24
(9H, s, 3 × CH3CO2), 4.68 (1H, m, w/2 = 12 Hz, H-17␣), 4.69 (1H,
dd, J = 9, 8 Hz, H-3␣), 4.98 (1H, t, J = 4.5 Hz, H-7␤), 5.59 (1H, d,
J = 5 Hz, H-6). The second transformed compound was 3␤,7␤,17␤-
trihydroxyandrost-5-ene (4). This product was characterized as
the triacetate 4a (22.7 mg), which crystallized as needles from
acetone-methanol, Rf = 0.88, acetone/dichloromethane (1:19 v/v),
m.p. 178–179◦, [␣]D +42.9◦ (c 0.15), lit [13]. 210–211◦, [␣]D +52◦;
bead fermentations, 17␤-acetoxyandrost-4-en-3-one (2a) (1.7 mg)
was also isolated.
2.5.4. Products from Mucor plumbeus and Cunninghamella
echinulata var. elegans
The extract was purified using column chromatography.
Elution with acetone/dichloromethane (1:9 v/v) afforded the
fed compound (1) (6.2 mg). Further purification and acetyla-
tion resulted in the recovery of three derivitized metabolites
that were found in the individual immobilized cell fermen-
tations [6]: 3␤,7␣,17␤-triacetoxyandrost-5-ene (3a) (7.6 mg),
3␤,7␤,17␤-triacetoxyandrost-5-ene (4a) (31.1 mg), and 3␤,7␣-
diacetoxyandrost-5-en-17-one (5a) (7.7 mg). Further purifications
using PLC gave eight metabolites that were not previously iso-
lated from the immobilized cell fermentation of either fungus.
The first metabolite was 7␣,17␤-dihydroxyandrost-4-en-3-one (9).
This was characterized as the diacetate 9a (17.9 mg), which resisted
crystallization, Rf = 0.62, acetone/dichloromethane (1:19 v/v), [␣]D
IR: ꢀmax 2947, 1739, 1733, 1370, 1238 cm−1 1H NMR: ␦ 0.79
;
(3H, s, H-18), 1.05 (3H, s, H-19), 2.10 (9H, s, 3 × CH3CO2), 4.60
(2H, d, J = 8.1 Hz, H-3␣, H-17␣), 5.08 (1H, d, J = 8.2 Hz, H-7␣), 5.37
(1H, s, H-6). The third product was 3␤,7␣-dihydroxyandrost-5-en-
17-one (5). This was characterized as the diacetate 5a (20.8 mg),
which resisted crystallization, Rf = 0.81, acetone/dichloromethane
(1:19 v/v), [␣]D −228◦ (c 0.21), lit [14]. m.p. 168–170◦, [␣]D −178◦;
IR: ꢀmax 1722, 1622, 1228 cm−1 1H NMR: ␦ 0.88 (3H, s, H-18), 1.05
;
(3H, s, H-19), 2.05 (6H, s, 2 × CH3CO2), 4.68 (1H, m, w/2 = 16.5 Hz,
H-3␣), 5.12 (1H, dt, J = 8.7, 2.1 Hz, H-7␤), 5.62 (1H, dd, J = 5.3, 1.5 Hz,
H-6). The final metabolite was 17␤-hydroxyandrost-4-en-3-one
(2). This was characterized as the acetate 2a (1.8 mg), which crystal-
lized from acetone as needles, Rf = 0.14, acetone/dichloromethane
(1:19 v/v), m.p. 130–132◦, [␣]D +92◦ (c 0.01), lit [15]. 141–142◦, [␣]D
+93.3◦ (c 0.18); IR: ꢀmax 2948, 1737, 1676, 1376, 1245 cm−1
;
HREIMS: m/z: 411.2162 [MNa]+ (C23H32O5Na requires 411.2141);
1H NMR: ␦ 0.84 (3H, s, H-18), 1.22 (3H, s, H-19), 2.03 (3H,
s, CH3CO2-7), 2.05 (3H, s, CH3CO2-17), 4.63 (1H, dd, J = 7.9,
9.0 Hz, H-17␣), 5.02 (1H, q, J = 3 Hz, H-7␣), 5.70 (1H, bs, H-4).
The second derivative was 3␤,14␣,17␤-trihydroxyandrost-5-en-
7-one (10). This was characterized as the diacetate 10a (3.2 mg),
which resisted crystallization, Rf = 0.34, acetone/dichloromethane
(1:19 v/v), [␣]D +102.7◦ (c 0.22); IR: ꢀmax 3630, 2924, 1735,
1374, 1245 cm−1; HREIMS: m/z: 427.2074 [MNa]+ (C23H36O6Na
requires 427.2091); 1H NMR: ␦ 0.89 (3H, s, H-18), 1.24 (3H,
s, H-19), 2.04 (3H, s, CH3CO2-3), 2.05 (3H, s, CH3CO2-17), 4.72
(1H, tt, J = 5, 11 Hz, H-3␣), 5.19 (1H, dd, J = 6.5, 9 Hz, H-17␣),
5.73 (1H, d, J = 1.5 Hz, H-6). The third transformed product was
3␤,5␤,6␣,7␣,17␤-pentahydroxyandrostane (11). This was charac-
terized as the triacetate 11a (3.8 mg), which resisted crystallization,
Rf = 0.12, acetone/dichloromethane (1:19 v/v), [␣]D +86.9◦ (c 0.26);
IR: ꢀmax 3649, 2936, 1733, 1375, 1244 cm−1; HREIMS: m/z (rel. int.):
489.2471 (100) [MNa]+ (C25H38O8Na requires 489.2458); 1H NMR:
␦ 0.79 (3H, s, H-18), 0.99 (3H, s, H-19), 2.05 (3H, s, CH3CO2-17), 2.09
(3H, s, CH3CO2-3), 2.10 (3H, s, CH3CO2-7), 3.85 (1H, d, J = 3.8 Hz,
H-6␤), 4.61 (1H, t, J = 8.5 Hz, H-17␣), 5.27 (1H, bs, H-3␣), 5.32
(1H, t, J = 3.6 Hz, H-7␤). The fourth product was 3␤,5␣,6␤,7␣,17␤-
pentahydroxyandrostane (12). This compound was characterized
as the tetraacetate 12a (1.8 mg), which resisted crystallization,
Rf = 0.28, acetone/dichloromethane(1:19v/v), [␣]D +163.8◦ (c 0.16),
IR: ꢀmax 3727, 2940, 1733, 1371, 1243 cm−1; HREIMS: m/z (rel. int.):
531.2562 (100) [MNa]+ (C27H40O9Na requires 531.2564); 1H NMR:
␦ 0.83 (3H, s, H-18), 1.17 (3H, s, H-19), 2.01 (3H, s, CH3CO2-7),
2.05 (3H, s, CH3CO2-17), 2.11 (3H, s, CH3CO2-6), 2.12 (CH3CO2-
3), 4.62 (1H, t, J = 8.5 Hz, H-17␣), 4.80 (1H, d, J = 2.3 Hz, H-6␣), 4.89
(1H, t, J = 3 Hz, H-7␤), 5.18 (1H, tt, J = 5.5, 11 Hz, H-3␣). The fifth
metabolite was 3␤,5␣,6␤,11␣,17␤-pentahydroxyandrostane (13).
This was characterized as the tetraacetate 13a (7.6 mg), which
resisted crystallization, Rf = 0.22, acetone/dichloromethane (1:19
+96.2◦; IR: ꢀmax 2932, 1737, 1675, 1247 cm−1 1H NMR: ␦ 0.84 (3H,
;
s, H-18), 1.20 (3H, s, H-19), 2.05 (3H, s, CH3CO2-17), 4.60 (1H, dd,
J = 7.7, 8.8 Hz, H-17␣), 5.74 (1H, s, H-4).
2.5.2. Products from Rhizopus oryzae and Mucor plumbeus
The extract (173 mg) was purified using column chromatog-
raphy. Elution with acetone/dichloromethane (1:9 v/v) afforded
the fed compound (1) (57.6 mg). Further purifications using
the same solvent system and acetylation resulted in the
recovery of acetylated derivatives of three compounds which
were previously isolated from the single bead fermentations:
3␤,7␣,17␤-triacetoxyandrost-5-ene (3a) (11.4 mg), 3␤,7␤,17␤-
triacetoxyandrost-5-ene (4a) (8 mg), and 3␤,7␣-diacetoxyandrost-
5-en-17-one (5a) (10 mg). Further purifications using PLC allowed
for the isolation of two compounds that were not previously found
in the single bead fermentations using either microorganism. The
first metabolite was 14␣,17␤-dihydroxyandrost-4-en-3-one (6).
This was characterized as the monoacetate 6a (3.9 mg), which
resisted crystallization, Rf = 0.19, acetone/dichloromethane (1:19
v/v), [␣]D +133.8◦ (c 0.16); IR: ꢀmax 3509, 2946, 1733, 1671,
1375, 1248 cm−1; HREIMS: m/z: 362.2149 M+ (C21H28O4 requires
362.2144); 1H NMR: ␦ 0.96 (3H, s, H-18), 1.24 (3H, s, H-19), 2.06
(3H, s, CH3CO2-17), 5.18 (1H, dd, J = 6.5, 9.0 Hz, H-17␣), 5.76 (1H,
bs, H-4). The second compound was 7␤,17␤-dihydroxyandrost-4-
en-3-one (7). This was characterized as the diacetate 7a (6.6 mg),
which resisted crystallization, Rf = 0.60, acetone/dichloromethane
(1:19 v/v), [␣]D +136.4◦ (c 0.28); IR: ꢀmax 2947, 1736, 1373, 1241,
1032 cm−1; HREIMS: m/z: 411.2160 [MNa]+ (C23H32O5Na requires
411.2141); 1H NMR: ␦ 0.86 (3H, s, H-18), 1.23 (3H, s, H-19), 2.04 (3H,
s, CH3CO2-7), 2.05 (3H, s, CH3CO2-17), 4.58 (1H, dd, J = 8.0, 9.0 Hz,
H-17␣), 4.63 (1H, td, J = 5, 10 Hz, H-7␣), 5.77 (1H, d, J = 1.9 Hz, H-4).
v/v), [␣]D +75.8◦ (c 0.24); IR: ꢀmax 3449, 2923, 1733, 1245 cm−1
;
2.5.3. Products from Rhizopus oryzae and Whetzelinia
sclerotiorum
The extract (214 mg) was purified by column chro-
matography. Elution with acetone/dichloromethane (1:9 v/v)
afforded the fed compound (1) (110 mg). Acetylation and
further purification yielded three derivitized metabo-
HREIMS: m/z (rel. int.): 531.2563 (100) [MNa]+ (C27H40O9Na
requires 531.2564); 1H NMR: ␦ 0.86 (3H, s, H-18), 1.27 (3H, s,
H-19), 1.97 (3H, s, CH3CO2-11), 2.02 (3H, s, CH3CO2-6), 2.03 (3H,
s, CH3CO2-17), 2.09 (CH3CO2-3), 4.64 (1H, t, J = 8.5 Hz, H-17␣),
4.68 (1H, bs, H-6␣), 5.09 (1H, m, w/2 = 12 Hz, H-3␣), 5.17 (1H, td,
J = 5.5, 10.5 Hz, H-11␤). The sixth congener was 3␤,5␣,6␤,7␤,17␤-
pentahydroxyandrostane (14). This was characterized as the
tetraacetate 14a (4.2 mg), which resisted crystallization, Rf = 0.12,
lites,
3␤,7␣,17␤-triacetoxyandrost-5-ene
(3a)
(51.4 mg),
3␤,7␤,17␤-triacetoxyandrost-5-ene (4a) (20.4 mg) and 3␤,7␤-
diacetoxyandrost-5-en-17-one (8a) (3 mg), that were derived from
compounds 3, 4 and 8 respectively, and all of which were previ-
ously isolated from the individual immobilized cell fermentations
[6]. A lone acylated metabolite, not found in the individual single
acetone/dichloromethane (1:19 v/v), [␣]D +75.4◦ (c 0.13); IR: ␯
max
3449, 2925, 1733, 1245 cm−1; HREIMS: m/z (rel. int.): 531.2562
(100) [MNa]+ (C27H40O9Na requires 531.2564); 1H NMR: ␦ 0.84
(3H, s, H-18), 1.19 (3H, s, H-19), 1.92 (3H, s, CH3CO2-6), 2.01

P.C. Peart et al. / Journal of Molecular Catalysis B: Enzymatic 125 (2016) 16–24
21
(3H, s, CH3CO2-7), 2.04 (3H, s, CH3CO2-17), 2.11 (CH3CO2-3), 4.56
2.6.2. Products from Mucor plumbeus and Cunninghamella
echinulata var. elegans
(1H, t, J = 8.5 Hz, H-17␣), 4.99 (1H, d, J = 4 Hz, H-6␣), 5.12 (1H,
m, w/2 = 22 Hz, H-3␣), 5.20 (1H, dd, J = 4, 11 Hz, H-7␣). The sev-
enth compound was 3␤,5␣,6␤,15␤,17␤-pentahydroxyandrostane
(15). This was characterized as the tetraacetate 15a (3.8 mg),
which resisted crystallization, Rf = 0.05, acetone/dichloromethane
The extract (200 mg) underwent partial purification by col-
umn chromatography. This allowed for the recovery of the
fed substrate (2) (14.6 mg). The first metabolite, isolated by
PLC using acetone/dichloromethane (1:49 v/v) (×3), was 17␤-
hydroxyandrosta-4,6-dien-3-one (21) (1.9 mg), which resisted
crystallization, Rf = 0.28, acetone/dichloromethane (1:19 v/v), [␣]D
+25.2◦ (c 0.14), lit [18,19]. m.p. 209−211◦, [␣]D +80◦; IR: ꢀmax
(1:19 v/v), [␣]D +160◦ (c 0.06); IR:
␯
max
3630, 2926, 1735,
1375, 1244 cm−1; HREIMS: m/z (rel. int.): 531.2569 (100) [MNa]+
(C27H40O9Na requires 531.2564); 1H NMR: ␦ 1.02 (3H, s, H-18),
1.21 (3H, s, H-19), 1.97 (3H, s, CH3CO2-15), 2.02 (3H, s, CH3CO2-
6), 2.04 (3H, s, CH3CO2-17), 2.05 (CH3CO2-3), 4.58 (1H, t, J = 8.5 Hz,
H-17␣), 4.65 (1H, t, J = 2.5 Hz, H-6␣), 4.96 (1H, ddd, J = 2, 5, 6 Hz,
H-15␣), 5.13 (1H, tt, J = 5, 10 Hz, H-3␣). The eighth derivative
was 3␤,5␣,6␤,14␣,17␤-pentahydroxyandrostane (16). This was
characterized as the triacetate 16a (3.4 mg), which resisted crystal-
lization, Rf = 0.03, acetone/dichloromethane (1:19 v/v), [␣]D +162◦
(c 0.08); IR: ꢀmax 3630, 2936, 1734, 1458, 1374 cm−1; HREIMS: m/z
(rel. int.): 489.2451 (100) [MNa]+ (C25H38O8Na requires 489.2458);
1H NMR: ␦ 0.92 (3H, s, H-18), 1.18 (3H, s, H-19), 2.01 (3H, s, CH3CO2-
6), 2.04 (3H, s, CH3CO2-17), 2.08 (3H, s, CH3CO2-3), 4.77 (1H, t,
J = 3 Hz, H-6␣), 5.12 (1H, tt, J = 5, 11 Hz, H-3␣), 5.16 (1H, dd, J = 9.0,
6.6 Hz, H-17␣).
3426, 2930, 1661, 1452, 1225 cm−1
;
1H NMR: ␦ 0.85 (3H, s,
H-18), 1.13 (3H, s, H-19), 3.70 (1H, t, J = 8.5 Hz, H-17␣), 5.68
(1H, s, H-4), 6.10 (1H, dd, J = 1.5, 4 Hz, H-7), 6.11 (1H, dd,
J = 1, 3 Hz, H-6). The second transformed product was puri-
fied by PLC using acetone/dichloromethane (1:49 v/v). This gave
7␣-acetoxy-17␤-hydroxyandrost-4-en-3-one (22) (3.2 mg), which
resisted crystallization, Rf = 0.18, acetone/dichloromethane (1:19
v/v), [␣]D +32.5◦ (c 0.08); IR: ꢀmax 3445, 1646, 1230 cm−1
;
HREIMS: m/z (rel. int.): 369.2046 (100) [MNa]+ (C21H30O4Na
requires 369.2036); 1H NMR: ␦ 0.80 (3H, s, H-18), 1.22 (3H,
s, H-19), 2.03 (3H, s, CH3CO2-7) 3.68 (1H, t, J = 8.5 Hz, H-17␣),
5.01 (1H, d, J = 2.8 Hz, H-7␤), 5.70 (1H, bs, H-4). The third con-
gener was 6␤-hydroxyandrost-4-ene-3,17-dione (17) (3 mg). The
fourth metabolite, purified via PLC using acetone/dichloromethane
(1:9 v/v) (× 3), was 7␣-hydroxyandrost-4-ene-3,17-dione (23)
(6.6 mg), which crystallized from acetone as needles, Rf = 0.26, ace-
tone/dichloromethane (1:19 v/v), m.p. 104–108◦, [␣]D +20.1◦ (c
2.6. Biotransformations using 17ˇ-hydroxyandrost-4-en-3-one
(2)
0.07), lit [20]. [␣]D +202◦; IR: ꢀmax 3447, 2944, 1653, 1223 cm−1
;
1H NMR: ␦ 0.92 (3H, s, H-18), 1.22 (3H, s, H-19), 4.09 (1H, m,
w/2 = 10 Hz, H-7␤), 5.83 (1H, s, H-4). The fifth biotransformed
compound, isolated with acetone/dichloromethane (1:9 v/v), was
6␤,17␤-dihydroxyandrost-4-en-3-one (19) (16.3 mg). The sixth
product of transformation, obtained in fractions that were
eluted with acetone/dichloromethane (3:17 v/v), was 7␤,17␤-
dihydroxyandrost-4-en-3-one (7) (9.9 mg), which crystallized from
dichloromethane as plates, Rf = 0.11, acetone/dichloromethane (1:9
v/v), m.p. 178−180◦, [␣]D +67.1◦ (c 0.07), lit [21]. m.p. 183.5–185.5◦,
2.6.1. Products from Rhizopus oryzae and Mucor plumbeus
Partial purification of the extract (204 mg) allowed for
the recovery of the fed substrate (2) (6.9 mg). Elution with
acetone/dichloromethane (1:19 v/v) afforded the first metabo-
lite, 6␤-hydroxyandrost-4-ene-3,17-dione (17) (25.8 mg), which
crystallized from acetone as amorphous crystals, Rf = 0.16, ace-
tone/dichloromethane (1:19 v/v), m.p. 80–81◦, [␣]D +46.3◦ (c 0.31),
lit [16]. m.p. 154–162◦; IR: ꢀmax 3429, 2360, 1733, 1679, 1230 cm−1
;
1H NMR: ␦ 0.93 (3H, s, H-18), 1.41 (3H, s, H-19), 4.41 (1H, t, J = 2.8 Hz,
H-6␣), 5.84 (1H, bs, H-4). The second compound was collected in
fractions that were eluted with acetone/dichloromethane (1:9 v/v).
11␣-Hydroxyandrost-4-ene-3,17-dione (18) (10.7 mg) crystallized
as needles from acetone, Rf = 0.19, acetone/dichloromethane (1:19
v/v), m.p. 182–184◦, [␣]D +68.6◦ (0.17), lit [17]. m.p. 240–241◦;
[␣]D +101◦; IR: ꢀmax 3395, 2941, 1640, 1435, 1233 cm−1 1H NMR:
;
␦ 0.82 (3H, s, H-18), 1.22 (3H, s, H-19), 3.63 (1H, t, J = 8.7 Hz,
H-17␣), 3.45 (1H, m, w/2 = 15 Hz, H-7␣), 5.77 (1H, s, H-4). Fur-
ther elution with acetone/dichloromethane (3:17 v/v) yielded the
seventh metabolite, 14␣,17␤-dihydroxyandrost-4-en-3-one (6)
(33.4 mg). The final product of transformation was also obtained
in fractions that were eluted from acetone/dichloromethane
(3:17 v/v). 7␣,17␤-Dihydroxyandrost-4-en-3-one (9) (19.5 mg)
crystallized from dichloromethane as plates, Rf = 0.12, ace-
tone/dichloromethane (3:17 v/v), m.p. 205–206◦, [␣]D +33.3◦ (c
0.42), lit [22]. m.p. 212–215◦, [␣]D +88◦; IR: ꢀmax 3411, 2941,
1661, 1232 cm−1; HREIMS: m/z: 304.2038 M+ (C19H28O3 requires
304.2038); 1H NMR: ␦ 0.79 (3H, s, H-18), 1.22 (3H, s, H-19), 3.65
(1H, t, J = 8.8 Hz, H-17␣), 3.94 (1H, d, J = 2.5 Hz, H-7␤), 5.78 (1H, s,
H-4).
IR: ꢀmax 3447, 2936, 1734, 1654, 1457 cm−1 1H NMR: ␦ 0.95
;
(3H, s, H-18), 1.35 (3H, s, H-19), 4.08 (1H, dt, J = 5.2, 10.5 Hz, H-
11␤), 5.76 (1H, s, H-4). The third metabolite was eluted with
acetone/dichloromethane (1:9 v/v). 6␤,17␤-Dihydroxyandrost-4-
en-3-one (19) (29.1 mg) crystallized as needles from acetone,
Rf = 0.14, 10% acetone in dichloromethane, m.p. 207–210◦, [␣]D
+9.41◦ (c 0.17), lit [16]. m.p. 206–210◦; IR: ꢀmax 3396, 2946,
1661, 1230 cm−1 1H NMR: ␦ 0.82 (3H, s, H-18), 1.39 (3H, s, H-
;
19), 3.66 (1H, t, J = 8.5 Hz, H-17␣), 4.39 (1H, t, J = 2.8 Hz, H-6␣),
5.85 (1H, bs, H-4). Further elution with acetone/dichloromethane
(1:9 v/v) allowed for the recovery of a fourth product, 14␣,17␤-
dihydroxyandrost-4-en-3-one (6) (15.9 mg), which crystallized
from acetone as amorphous crystals, Rf = 0.10, 10% acetone in
dichloromethane (1:9 v/v), m.p. 183−188◦, [␣]D +77.9◦ (c 0.78),
lit [16]. m.p. 181–184◦, [␣]D +121.8◦; IR: ꢀmax 3425, 2944, 1709,
3. Results and discussion
3.1. Biotransformations using 3ˇ,17ˇ-dihydroxyandrost-5-ene
(1)
1660 cm−1 1H NMR: ␦ 0.90 (3H, s, H-18), 1.22 (3H, s, H-19), 4.25
;
(1H, dd, J = 9.0, 6.8 Hz, H-17␣), 5.73 (1H, bs, H-4). PLC purifica-
tion using acetone/dichloromethane (3:17 v/v) yielded the final
transformed product, 11␣,17␤-dihydroxyandrost-4-en-3-one (20)
(20.5 mg), which crystallized as cubes from acetone, Rf = 0.08, ace-
tone/dichloromethane (1:9 v/v), m.p. 168–172◦, [␣]D +36.5◦ (c
1.13), lit [17]. m.p. 218–219◦; IR: ꢀmax 3397, 2938, 1657, 1356,
3.1.1. Products from Rhizopus oryzae and Cunninghamella
echinulata var. elegans
The incubation of 3␤,17␤-dihydroxyandrost-5-ene (1) with
immobilized cells of R. oryzae and C. echinulata var. elegans
yielded four metabolites, two of which were found in the
single bead fermentations: 3␤,7␣,17␤-trihydroxyandrost-5-ene
(3) and 3␤,7␤,17␤-trihydroxyandrost-5-ene (4) [6]. The other
compounds, 3␤,7␣-dihydroxyandrost-5-en-17-one (5) and 17␤-
1238 cm−1 1H NMR: ␦ 0.81 (3H, s, H-18), 1.34 (3H, s, H-19), 3.69
;
(1H, t, J = 8.6 Hz, H-17␣), 3.98 (1H, td, J = 10.5, 4.8 Hz, H-11␤), 5.73
(1H, bs, H-4).

22
P.C. Peart et al. / Journal of Molecular Catalysis B: Enzymatic 125 (2016) 16–24
hydroxyandrost-4-en-3-one (2), were not found in the single bead
fermentation using either fungus. These compounds were charac-
terized as their acetates (3a, 4a, 5a and 2a respectively). It is likely
that steroid 5 was generated by oxidation of the C-17 alcohol of 3
by one of the fungi. The formation of 2 possibly occurred via oxi-
dation of the 3-hydroxyl of 1, followed by rearrangement of the
5,6-double bond to give the conjugated enone.
resonance at ␦ 70.3 was approximately 5 ppm upfield of that
in compound 7a. The T-ROESY data showed that H-7 was cou-
pled to H-8 and H-15␤, indicating that the stereochemistry of the
hydroxyl group was ␣. Hence, this compound was deduced to be
7␣,17␤-diacetoxyandrost-4-en-3-one (9a). This was derived from
7␣,17␤-dihydroxyandrost-4-en-3-one (9). Steroid 9 was probably
formed by 7␣-hydroxylation of 2, the generation of which was pro-
posed above. The 13C NMR spectrum of compound 10a revealed
the appearance of a new nonprotonated carbon at ␦ 201.8. This
pointed towards the presence of a conjugated carbonyl; however,
the alcohol at C-3 remained, and this was a clear indication that
the conjugated ketone was not located in ring A. The HMBC data
showed that H-8 was coupled to this new carbonyl; therefore, this
ketone moiety had to be located at C-7. Additionally, there was a
new nonprotonated carbon, resonating at ␦ 80.4. This was indica-
tive of oxygen insertion at a methine position. The HMBC data also
placed this new alcohol in ring D, as there was coupling between
H-18 and the nonprotonated carbon signal. It was deduced that
this compound was 3␤,17␤-diacetoxy-14␣-hydroxyandrost-5-en-
7-one (10a), which was derived from 10. The latter would have
been formed by 14␣-hydroxylation of 1. The NMR data of the tri-
acetate (11a) of the third metabolite (11) displayed the loss of the
C-5,6 double bond. This was accompanied by the appearance of
three new functionalized carbons at ␦ 71.4, 72.7 and 76.7. The HSQC
spectrum showed that the methine at ␦ 71.4 was bonded to a pro-
ton resonating at ␦ 5.32. This proton signal was correlated to H-8 in
the COSY spectrum. Consequently, the site of oxygen insertion was
located at C-7. The T-ROESY spectrum showed couplings between
H-7 and H-6␤, H-8, H-15␤, and as such the stereochemistry of the
acetoxy group was determined to be ␣. From the HSQC data it was
extrapolated that the resonance at ␦ 76.7 represented the carbon
of a tertiary alcohol. It was determined that this hydroxyl moi-
ety was located at C-5. The T-ROESY data showed no cross peaks
between H-4 and H-19. It was determined that this was due to
the presence of a cis-fused ring junction. Therefore, the stereo-
chemistry of the C-5 hydroxyl must have been ␤. Subsequently,
the methine at ␦ 72.7 was proposed to be located at C-6. This was
supported by the HMBC data that showed the corresponding pro-
ton to be coupled to C-4, C-5 and C-7. T-ROESY data also revealed
that H-6 was coupled to H-7␤, H-8 and H-19. As such the stereo-
chemistry of the hydroxyl group was deduced to be ␣. Compound
11a was 3␤,7␣,17␤-triacetoxy-5␤,6␣-dihydroxyandrostane, a pre-
viously unreported compound, which was derived from compound
11. The fourth metabolite (12) was characterized the acetate (12a).
The 13C NMR of the latter showed similarities to that of 11a. The
C-5,6 double bond was no longer present, and there were three
new functionalized carbons. There was a new nonprotonated car-
bon signal at ␦ 75.8, which was deduced to be the resonance for C-5.
The COSY data showed that the signals at ␦ 71.6 (13C) and ␦ 4.89
(1H) were coupled to H-6 and H-8. Therefore, this acetoxy group
was located at C-7. The HSQC spectrum indicated that the carbon
with resonance at ␦ 73.0 was bonded to the proton at ␦ 4.80. In
the COSY spectrum coupling was observed between this proton
and H-7. It was inferred that this methine was at C-6. This was
further supported by the HMBC data that showed that H-6 corre-
lated to C-5, C-7, C-8, and C-10. This compound was deduced to be
3␤,6␤,7␣,17␤-tetraacetoxy-5␣-hydroxyandrostane (12a), which
was derived from 3␤,5␣,6␤,7␣,17␤-pentahydroxyandrostane (12).
Analogues 11 and 12 are probably formed by 7␣-hydroxylation
of 1 to give 3, that is then epoxidized. Enzyme-catalyzed hydra-
tion of the resulting 5,6-epoxide could lead to either 11 or 12. The
spectral data of compound 13a, derived from the fifth metabolite
(13), showed similarities to those of the preceding compounds.
The C-5,6 double bond was no longer present. The 13C NMR
spectrum displayed three new resonances which corresponded to
functionalized carbons. The signal at ␦ 75.7 represented a non-
3.1.2. Products from Rhizopus oryzae and Mucor plumbeus
This fermentation afforded five metabolites, three of which
were also isolated from the single bead incubations: 3␤,7␣,17␤-
trihydroxyandrost-5-ene (3), 3␤,7␤,17␤-trihydroxyandrost-5-ene
(4) and 3␤,7␣-dihydroxyandrost-5-en-17-one (5) [6]. The 13C NMR
data for the monoacetate (6a) of one of the ‘new’ metabolites (6)
showed the appearance of a nonprotonated carbon at ␦ 199.4,
which was indicative of the presence of a conjugated carbonyl
moiety. There was a significant downfield shift of 30.1 ppm in the
resonance value for C-5. This pointed towards a migration of the
carbon–carbon double bond from C-5,6 to C-4,5, after oxidation
of the C-3 alcohol. Further inspection of the NMR data revealed
that there was also another nonprotonated carbon present at ␦
82.9. The HMBC spectrum showed couplings between H-12, H-
18 and this new resonance, effectively placing the site of oxygen
insertion in ring D. The only position that met these requirements
was C-14. This compound was determined to be 17␤-acetoxy-
14␣-hydroxyandrost-4-en-3-one (6a), which was derived from
14␣,17␤-dihydroxyandrost-4-en-3-one (6) [6]. Oxidation of 1 with
double bond rearrangement to 2 (as seen above), followed by 14␣-
hydroxylation could be a possible route to 6. The NMR data for the
diacetate (7a) of the final product of transformation (7) showed a
similar feature to that of compound 6a, where there was a new
nonprotonated carbon present at ␦ 198.7, pointing towards the
presence of a conjugated ketone. There was also a new methine that
resonated at ␦ 75.5. The COSY data showed correlations between
H-8 and the new methine proton, indicating that oxygen insertion
had occurred at C-7. This was further corroborated by the HMBC
data that showed couplings between H-5, H-6, H-8, H-9 and this
new methine. T-ROESY data indicated that H-7 was correlated to
H-6␣, H-9 and H-14; hence the stereochemistry of the alcohol was
determined to be ␤. This analogue was deduced to be 7␤,17␤-
diacetoxyandrost-4-en-3-one (7a), which was derived from 7 [6].
Compound 7 is envisaged as the 7␤-hydroxylation product of 2.
3.1.3. Products from Rhizopus oryzae and Whetzelinia
sclerotiorum
This biotransformation afforded four metabolites, three of
which were previously isolated from the individual immobi-
lized cell fermentations: 3␤,7␣,17␤-trihydroxyandrost-5-ene
(3), 3␤,7␤,17␤-trihydroxyandrost-5-ene (4) and 3␤,7␤-
dihydroxyandrost-5-en-17-one (8). These compounds were
characterized as their acetates (3a, 4a, and 8a) respectively [6]. The
sole metabolite which was not previously found in the single bead
fermentation was 17␤-hydroxyandrost-4-en-3-one (2), character-
ized as its acetate (2a). The proposed path to this compound has
been outlined above.
3.1.4. Products from Mucor plumbeus and Cunninghamella
echinulata var. elegans
Eleven metabolites were isolated, eight of which were not
found in the single fungal bead fermentations. The three congeners
that were previously isolated were 3␤,7␣,17␤-trihydroxyandrost-
5-ene (3), 3␤,7␤,17␤-trihydroxyandrost-5-ene (4) and 3␤,7␣-
dihydroxyandrost-5-en-17-one (5), all characterized as their
respective acetates (3a–5a) [6]. The first “new” metabolite (9)
was characterized as the acetate (9a). The NMR data of the latter
was similar to that of compound 7a. However, the new methine

P.C. Peart et al. / Journal of Molecular Catalysis B: Enzymatic 125 (2016) 16–24
23
protonated carbon, and as such was assigned to the C-5 position.
Based on information obtained from the HSQC and COSY spec-
tra it was deduced that the carbon at ␦ 76.0 was located at C-6.
The T-ROESY data revealed that H-6 was correlated to H-4␣. This
indicated that the stereochemistry of the acetoxyl group was ␤.
The carbon (methine) signal at ␦ 70.3 was attached to the pro-
ton at ␦ 5.11. The COSY spectrum showed that the latter was
coupled to H-9 and H-12. Hence, the position of hydroxylation
was determined to have been at C-11. The T-ROESY spectrum
showed H-11 was correlated to H-18 and H-19, and as such the
stereochemistry of the acetoxy group was determined to be ␣.
This compound was designated to be the novel 3␤,6␤,11␣,17␤-
tetraacetoxy-5␣-hydroxyandrostane (13a), which was derived
from 3␤,5␣,6␤,11␣,17␤-pentahydroxyandrostane (13). The forma-
tion of 13 is imagined to involve 11␣-hydroxylation of 1. The
resulting triol is then epoxidized and hydrated as described pre-
viously to yield 13. The tetraacetate (14a) of the sixth metabolite
(14) exhibited the same features as the preceding congeners in the
NMR spectra. The double bond was no longer present and there
were three new carbons bearing oxygen funtionalities. The ace-
toxy bearing carbon with resonance at ␦ 75.9 was nonprotonated,
and was determined to be located at C-5. The methine at ␦ 5.02
and ␦ 73.0 was coupled to H-6 and H-8, which indicated it was
located at C-7. The proton attached to the methine carbon at ␦
74.9 showed couplings to C-5, C-7 and C-8 in the HMBC spectrum;
therefore, this methine was located at C-6. The coupling constant
for H-6 was 4 Hz. Based on this the proton would be equatorial
(␣). Thus, the stereochemistry of the acetoxyl group must be ␤. H-
7 appeared as a doublet of doublets with couplings constants of 4,
11 Hz. Based on these values it was determined that this proton was
axial. The stereochemistry of H-7 was deduced to be ␣ and as such
the acetoxyl group must be ␤. This compound was determined to be
3␤,6␤,7␤,17␤-tetraacetoxy-5␣-hydroxyandrostane (14a), which
was derived from compound (14). There are no previous reports
in the literature of this compound. Steroid 14 is probably formed
from 4, that is then epoxidized and hydrated. As was found with the
previous metabolites the tetraacetate of seventh transformed prod-
uct (15a) no longer had a carbon–carbon double bond present, and
there were three new functionalized carbons. The nonprotonated
carbon bearing a hydroxyl and resonating at ␦ 75.9 was denoted as
C-5. The HSQC and COSY data revealed that the methine at ␦ 76.0
was located at C-6. The T-ROESY data showed that H-6 was coupled
to H-4␣ and H-7␣. Therefore, the stereochemistry of the acetoxy
group was ␤. The proton of the final methine was coupled to H-
14 and H-16, based on the information from the COSY spectrum,
and therefore the position of hydroxylation must have been C-15.
The T-ROESY data was used to establish the stereochemistry as
being ␤. Correlations were observed between H-14, H-16␣ and H-
15. Compound 15a was determined to be the hitherto unreported
3␤,6␤,15␤,17␤-tetraacetoxy-5␣-hydroxyandrostane, which was
derived from 3␤,5␣,6␤,15␤,17␤-pentahydroxyandrostane (15).
The latter could be formed firstly by hydroxylation at C-15␤, fol-
lowed by epoxidation and hydration. For the triacetate (16a) of the
final product of transformation (16) it was observed, as with pre-
vious analogues, that the double bond was absent. However, the
compound possessed two tertiary alcohols. The HMBC data showed
that the methine resonating at ␦ 74.9 was coupled to H-4, H-6 and
H-19. This indicated that this carbon was located between rings A
and B, and therefore it was reasoned to be at C-5. The second non-
protonated oxygen bearing carbon was correlated to H-18; hence,
the site of oxygen insertion was at C-14. The final site of hydrox-
ylation was determined to be at C-6. This was supported by the
coupling observed between the proton and H-7 (COSY data). H-6
appeared as a triplet with a coupling constant of 3 Hz. This indicated
that this proton was equatorial, therefore, the stereochemistry of
the acetoxyl moiety was ␤. This compound was determined to be
3␤,6␤,17␤-triacetoxy-5␣,14␣-dihydroxyandrostane (16a), which
was derived from compound 16. This compound was not previously
reported in the literature. Its formation would possibly involve the
generation of 11 and that is sequentially epoxidized and hydrated.
3.2. Biotransformations using 17ˇ-hydroxyandrost-4-en-3-one
(2)
3.2.1. Products from Rhizopus oryzae and Mucor plumbeus
This incubation afforded five metabolites, four of
which were isolated from the single bead fermentations:
14␣,17␤-dihydroxyandrost-4-en-3-one (6), 6␤-hydroxyandrost-
4-ene-3,17-dione (17), 6␤,17␤-dihydroxyandrost-4-en-3-one
(19) and 11␣,17␤-dihydroxyandrost-4-en-3-one (20) [7]. The
13C NMR for the lone ‘new’ metabolite showed the appearance
of a nonprotonated carbon, resonating at ␦ 218.4. This change
was accompanied by the loss of the methine at ␦ 81.9, which
pointed towards oxidation of the C-17 alcohol. This was supported
by HMBC correlations between the carbon at ␦ 218.4 and H-12,
H-15, H-16 and H-18. In the 1H NMR spectrum there was a new
resonance at ␦ 4.08, which was coupled to H-9 and H-12 (COSY).
From this it was interpreted that a new methine was located at
C-11. The T-ROESY data was used to establish the stereochemistry
of the hydroxyl as ␣, due to the correlations observed between
H-16␤, H-18, H-19 and H-11␤. This congener was determined to
be 11␣-hydroxyandrost-4-ene-3,17-dione (18). This would have
been formed by oxidation of 20 at C-17.
3.2.2. Products from Mucor plumbeus and Cunninghamella
echinulata var. elegans
This bioconversion yielded eight metabolites, four
of which were found in the single bead fermentations:
14␣,17␤-dihydroxyandrost-4-en-3-one (6), 6␤-hydroxyandrost-
4-ene-3,17-dione (17), 6␤,17␤-dihydroxyandrost-4-en-3-one (19)
and 7␣,17␤-dihydroxyandrost-4-en-3-one (9) [7]. The 1H NMR
spectrum of the first ‘new’ product (21) revealed the appearance
of two new protons in the olefinic region at ␦ 6.10 and 6.11. The
COSY data showed that they were coupled to each other, while
the proton at ␦ 6.10 was also coupled to H-8. Therefore, it was
determined that a double bond was located between C-6 and C-7.
There were no other significant changes in the NMR data and as
such compound 21 was deduced to be 17␤-hydroxyandrosta-4,6-
dien-3-one. This steroid may have been formed from 17, which
was then dehydrated. 1H NMR data for the second compound
(22) revealed a new singlet with resonance at ␦ 2.08 which was
bonded to the carbon at ␦ 21.2 (HSQC data). The DEPT experiments
revealed that this was a new methyl and its resonance frequency
suggested that it was part of an acetyl group. This was supported
by the presence of a new nonprotonated carbon resonating at ␦
170.6. Further inspection of the NMR spectra showed that there
was a new methine at ␦ 70.5 (13C) and ␦ 5.01 (1H). This proton was
coupled with H-6 and H-8, and this indicated that oxygen insertion
had taken place at C-7, followed by biological acetylation. H-7
was correlated to H-6␤, H-8, H-15␤ and H-19, indicating that the
stereochemistry of the acetoxy group was ␣. This analogue was
determined to be 7␣-acetoxy-17␤-hydroxyandrost-4-en-3-one
(22). Compound 22 is envisaged to be formed by acetylation of
9. The 13C NMR data of the third compound (23) showed the
loss of the methine signal at ␦ 81.9, which was accompanied by
the appearance of a new nonprotonated carbon at ␦ 220.1. This
suggested that there was oxidation of the C-17 hydroxyl group.
There was a new proton at ␦ 4.09, which was coupled to H-6 and
H-8 in the COSY spectrum. This indicated that a new alcohol was
located at C-7. The T-ROESY data revealed that H-7 was coupled
H-1␤, H-6␤, H-8 and H-15␤. Therefore, the stereochemistry of the
hydroxyl was determined to be ␣. This metabolite was deduced

24
P.C. Peart et al. / Journal of Molecular Catalysis B: Enzymatic 125 (2016) 16–24
to be 7␣-hydroxyandrost-4-ene-3,17-dione (23). This compound
probably is formed by oxidation of 9 at C-17. The spectral data of
the final new metabolite (7) was quite similar to that of compound
23, however, the resonance value for the new methine showed a
significant difference of 12.6 ppm. The proton of this methine was
coupled to H-6 and H-8 in the COSY spectrum indicating that the
site of hydroxylation was C-7. The T-ROESY data showed that H-7
was correlated to H-6␣, H-9 and H-14. Consequently, the stereo-
chemistry of this hydroxyl moiety must have been ␤. This analogue
was elucidated to be 7␤,17␤-dihydroxyandrost-4-en-3-one (7). It
is possible that 7 comes from 9 via redox chemistry at C-7.
Acknowledgments
The authors wish to thank the University of the West Indies,
Mona/Inter-American Development Bank (UWI/IDB) Programme
and the UWI Board for Graduate Studies & Research for funding.
PCP is grateful to Tanaud International B.V. (Jamaica) for the grant-
ing of a postgraduate scholarship, and to the University of the West
Indies for providing a Tutorial Assistantship.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.molcatb.2015.11.
026.
4. Conclusion
The results show that the mixture of beads derived from two
different fungi can effect transformation of steroid substrates to
yield compounds not seen in the fermentation utilizing immobi-
lized cells from either organism. In this study the aim was not
to optimize product yield, but to determine if crossover occurred.
Therefore, fermentations were not run until the fed substrate
was completely consumed. There is great potential in employing
this method for generation of small libraries of compounds. This
was particularly evident in the case of the fermentation using M.
plumbeus and C. echinulata var. elegans for the bioconversion of
3␤,17␤-dihydroxyandrost-5-ene (1), where eight metabolites, not
present in the individual fermentations were observed. Seven of
these were novel compounds. Similarly, when 2 was incubated with
beads derived from M. plumbeus and C. echinulata var. elegans, four
additional metabolites were formed, one not previously reported
in the literature. The fact that supplementary compounds were
formed in the mixed fermentations suggests that in some cases
the substrate is first functionalized by one microorganism and the
product thus formed is further processed by the second fungus.
In other instances the mechanisms operating are not as obvious.
It is not easy to elucidate the pathways by which the products of
crossover are formed. Experiments in which products from a fer-
mentation using beads of fungus A are then incubated with fungus
B might provide some insight into what is happening; particularly
if this is followed by an examination of the results when biotrans-
formed compounds from encapsulated fungus B are then incubated
with fungus A. Extensions of this work could involve mixtures of
beads generated from three or more microorganisms. Moreover,
preliminary results in our laboratory show that this method has
application in terpene transformation as well.
References
[1] A.M. O’Reilly, J.A. Scott, Enzym. Microb. Technol. 17 (1995) 636–646.
[2] S. Kundu, A.C. Mahapatra, V. Kumar Nigam, K. Kundu, Artif. Cells Blood
Substit. Biotechnol. 31 (2003) 313–327.
[3] Y.H. Khang, H. Shankar, F. Senatora, Biotechnol. Lett. 10 (1988) 867–872.
[4] P. Wikström, E. Szwajcer, P. Brodelius, K. Nilsson, K. Mosbach, Biotechnol.
Lett. 4 (1982) 153–158.
[5] P. Adlercreutz, O. Holst, B. Mattiasson, Enzym. Microb. Technol. 4 (1982)
332–336.
[6] P.C. Peart, A.R.M. Chen, W.F. Reynolds, P.B. Reese, Steroids 77 (2012) 85–90.
[7] P.C. Peart, W.F. Reynolds, P.B. Reese, J. Mol. Catal. B: Enzym. 95 (2013) 70–81.
[8] C.D. Hufford, F.A. Badria, M. Abou-Karram, W.T. Shier, R.D. Rogers, J. Nat. Prod.
54 (1991) 1543–1552.
[9] A.S. Lamm, A.R.M. Chen, W.F. Reynolds, P.B. Reese, Steroids 72 (2007)
713–722.
[10] A. Belan, J. Bolte, A. Fauvre, J.G. Gourcy, H. Veschambre, J. Org. Chem. 52
(1987) 256–260.
[11] A.R.M. Chen, P.L.D. Ruddock, A.S. Lamm, W.F. Reynolds, P.B. Reese,
Phytochemistry 66 (2005) 1898–1902.
[12] M.R. Wilson, W.A. Gallimore, P.B. Reese, Steroids 64 (1999) 834–843.
[13] G. Defaye, M.J. Luche, E.M. Chambaz, J. Steroid Biochem. 9 (1978) 331–336.
[14] R.M. Dodson, R.T. Nicholson, R.D. Muir, J. Am. Chem. Soc. 81 (1959)
6295–6297.
[15] H. Tada, Y.K. Sawa, J. Org. Chem. 33 (1968) 3347–3350.
[16] S.-H. Hu, G. Genain, R. Azerad, Steroids 60 (1995) 337–352.
[17] M.I. Choudhary, S. Sultan, M.T.H. Khan, A. Yasin, F. Shaheen, Atta-ur-Rahman,
Nat. Prod. Res. 18 (2004) 529–535.
[18] A. Wettstein, Helv. Chim. Acta 23 (1940) 388–399.
[19] R.E. Schaub, M.J. Weiss, J. Org. Chem. 26 (1961) 3915–3925.
[20] F. Roussi, Q.A. Ngo, S. Thoret, F. Guéritte, D. Guénard, Eur. J. Org. Chem. 18
(2005) 3952–3961.
[21] R.M. Dodson, S. Mizuba, J. Org. Chem. 27 (1962) 698–700.
[22] K. Irmscher, W. Beerstecher, H. Metz, R. Watzel, K.-H. Bork, Chem. Ber. 97
(1964) 3363–3373.