Efficient hydroxylation of functionalized steroids by Colletotrichum lini ST-1

Article abstract of DOI:10.1016/j.molcatb.2015.07.003

Abstract Biotransformation of a series of steroid compounds (estradiol, estrone, androst-4-en-3,17-dione, testosterone, canrenone, 16α,17α-epoxyprogesterone and progesterone) with Colletotrichum lini ST-1 as biocatalyst was investigated. With the exception of estradiol, estrone and progesterone, the microorganism could selectively hydroxylate steroid substrates (4 g/L) with 70-85% conversion rate and 60-76% total products yield. The different hydroxylation sites between androst-4-en-3,17-dione (3) and testosterone (4) suggested that the hydroxyl group or carbonyl group on the substrate at C17 had profound influence on the location of introduced hydroxyl groups. Transformations of 3-keto-steroid (3, 4, 5, 6 and 7) included monohydroxylation or dihydroxylation at 11α and 15α positions, while hydroxylations of 3-hydroxy-steroid (DHEA) were hydroxylation at 7α and 15α positions. Moreover, time course experiments demonstrated dihydroxylation of androst-4-en-3,17-dione (3), canrenone (5) and 16α,17α-epoxyprogesterone (6) were all initiated by hydroxylation on ring-D (C15) followed by attack on ring-C (C11). In this study, several new hydroxylation products were discovered, including 11α,15α-dihydroxyandrost-4-en-3,17-dione (9), 11α,15α-dihydroxy-canrenone (12) and 11α,15α-dihydroxy-16α,17α-epoxyprogesterone (14). The breadth of substrate spectrum and the excellent conversion rates achieved with this fungus indicated that C. lini ST-1 was a potential microorganism for production of valuable pharmaceutical ingredients and precursors.

Full text of DOI:10.1016/j.molcatb.2015.07.003

Journal of Molecular Catalysis B: Enzymatic 120 (2015) 111–118
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Efficient hydroxylation of functionalized steroids by Colletotrichum
lini ST-1
Yan Wua,b, Hui Lia, Xiao-Mei Zhanga, Jin-Song Gonga, Zhi-Ming Raob, Jin-Song Shia,
Xiao-Juan Zhanga, Zheng-Hong Xua,∗
a School of Pharmaceutical Science, Jiangnan University, Wuxi 214122, People’s Republic of China
b The Key Laboratory of Industrial Biotechnology, Ministry of Education, Jiangnan University, Wuxi, People’s Republic of China
a r t i c l e i n f o
a b s t r a c t
Article history:
Biotransformation of a series of steroid compounds (estradiol, estrone, androst-4-en-3,17-dione, testos-
terone, canrenone, 16␣,17␣-epoxyprogesterone and progesterone) with Colletotrichum lini ST-1 as
biocatalyst was investigated. With the exception of estradiol, estrone and progesterone, the microor-
ganism could selectively hydroxylate steroid substrates (4 g/L) with 70–85% conversion rate and
60–76% total products yield. The different hydroxylation sites between androst-4-en-3,17-dione (3)
and testosterone (4) suggested that the hydroxyl group or carbonyl group on the substrate at C17
had profound influence on the location of introduced hydroxyl groups. Transformations of 3-keto-
steroid (3, 4, 5, 6 and 7) included monohydroxylation or dihydroxylation at 11␣ and 15␣ positions,
while hydroxylations of 3-hydroxy-steroid (DHEA) were hydroxylation at 7␣ and 15␣ positions.
Moreover, time course experiments demonstrated dihydroxylation of androst-4-en-3,17-dione (3), can-
renone (5) and 16␣,17␣-epoxyprogesterone (6) were all initiated by hydroxylation on ring-D (C15)
followed by attack on ring-C (C11). In this study, several new hydroxylation products were discov-
ered, including 11␣,15␣-dihydroxyandrost-4-en-3,17-dione (9), 11␣,15␣-dihydroxy-canrenone (12)
and 11␣,15␣-dihydroxy-16␣,17␣-epoxyprogesterone (14). The breadth of substrate spectrum and the
excellent conversion rates achieved with this fungus indicated that C. lini ST-1 was a potential microor-
ganism for production of valuable pharmaceutical ingredients and precursors.
Received 23 January 2015
Received in revised form 7 July 2015
Accepted 7 July 2015
Available online 10 July 2015
Keywords:
Colletotrichum lini
Steroids
Hydroxylation
Biotransformation
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
Numerous microbial bioconversions of steroids and sterols have
been reported and this number is tending to grow [6–10]. Most
Steroids with unique biological activity are widely used as anti-
inflammatory, diuretic, anabolic, contraceptive, antiandrogenic,
progestational, and anticancer agents [1]. Steroid pharmaceuticals
are ranked among the most marketed medical products and rep-
resent the second largest category next to antibiotics [2]. Given
the complex structure of the steroid molecule, the traditional
chemical synthesis of steroid compounds requires complicated and
multi-step schemes, which makes the process expensive, time con-
suming and causes serious problems in environmental disposal
[3,4]. Microbial transformation can be performed in mild conditions
with high regio- and stereo-selectivity, which is hardly available
for chemical synthesis [5]. Therefore, microbial transformation has
been widely used in the conversion and biosynthesis of steroid
pharmaceuticals.
of these reports were focusing mainly on steroid hydroxylations,
1-dehydrogenation and sterol side-chain cleavage. Among them,
hydroxylations are possibly the most applied type of steroids func-
tionalization. Compared with their less polar non-hydroxylated
analogs, steroids with hydroxyl group often possess a higher bio-
logical activity which is beneficial for therapeutic applications.
Current trend in steroid hydroxylation is devoted to searching
novel biocatalysts capable of performing reactions at positions C7␣,
C9␣, C11␣, C11␤, C15␣, and C16␣, which are mainly applied for
the production of pharmaceutical ingredients and their precursors
[1,11–16]. Although research developments have discovered many
potentially useful strains and obtained various steroid compounds,
efforts are ongoing to isolate microorganisms capable of perform-
ing the required steroid transformations or synthesizing of novel
steroid compounds.
Over 1500 species, most of which belong to the genus of
Mycobacterium, Streptomyces, Curvularia, Colletotrichum, Aspergillus
and Penicillium have been studied in the context of their application
∗
Corresponding author.
E-mail address: zhenghxu@jiangnan.edu.cn (Z.-H. Xu).
http://dx.doi.org/10.1016/j.molcatb.2015.07.003
1381-1177/© 2015 Elsevier B.V. All rights reserved.

112
Y. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 120 (2015) 111–118
Fig. 1. Transformation of DHEA (16) by C. lini ST-1.
in the conversion of steroids [17–21]. The strain Colletotrichum lini
ST-1, as we mentioned in previous study [22], possessed an active
monooxygenase which could transform DHEA (10 g/L) into 7␣-OH-
DHEA and 7␣,15␣-diOH-DHEA (Fig. 1) with high conversion rate.
Therefore, in order to broaden C. lini ST-1 substrate spectrum and
obtain more variety of valuable intermediates, as well as to explore
the underlying hydroxylation mechanism between reaction sites
and substrate structures, biotransformation of more steroid com-
pounds by this microorganism have to be further studied.
In this paper, a range of structurally diverse compounds
including estranes [estradiol (1), estrone (2)], androstanes
[androst-4-en-3,17-dione (3), testosterone (4)], and pregnanes
[canrenone (5), 16␣,17␣-epoxyprogesterone (6), progesterone (7)]
were incubated with C. lini ST-1. Then the transformed products
and sequence of hydroxylations were further determined.
concentrated in vacuum. Finally, 1022–1108 mg of crude extract
was obtained.
2.5. Analysis of transformation products
The crude extracts obtained from the biotransformation of vari-
ous substrates were separated by column chromatography on silica
gel using petroleum ether-ethyl acetate (5:5) to afford 8 and 9,
petroleum ether-ethyl acetate (4:6) to afford 10, dichloromethane-
methanol (15:1) to afford 11 and 12, dichloromethane-methanol
(12:1) to afford 13 and 14, and dichloromethane-methanol (97:3)
to afford 15. Compounds were analyzed by TLC using solvent
system I (chloroform-methanol, 98:2) to separate the trans-
formation sample of androst-4-en-3,17-dione, solvent system II
(chloroform-methanol, 95:5) to separate the transformation sam-
ple of testosterone, solvent system III (chloroform-methanol, 9:1)
to separate the transformation samples of canrenone and pro-
gesterone, solvent system IV (chloroform-methanol, 85:15) to
separate transformation sample of 16␣,17␣-epoxyprogesterone,
and all compounds were visualized by spraying with a solution
of sulfuric acid/ethanol (1:9) followed by heating at 105 ◦C for
5 min.
2. Material and methods
2.1. Chemicals and reagents
All steroid substrates were purchased from Tianjin Pharmaceu-
tical Company. Glucose, yeast extract and other chemicals were
obtained from commercial sources. Acetonitrile, ethyl acetate and
other solvents were of analytical grade. Merck 9385 silica gel was
used for column chromatography.
All substrates and transformation products were analyzed by
HPLC which was performed using Dionex HPLC (USA) with an Agi-
lent C18 column (5.0 ␮m, 4.6 mm × 250 mm) at the wavelength of
206 nm, column temperature of 30 ◦C, with gradient concentration
of acetonitrile/water (30:70 to 80:20) as the mobile phase at a flow
rate of 0.5 mL/min.
2.2. Microorganism
C. lini ST-1 was isolated previously in our laboratory and main-
tained at 4 ◦C on slant consisting of the following composition (g/L):
glucose 15, FeSO4 0.01, NaNO3 3, K2HPO4 1, KCl 0.5, MgSO4·7H2O
0.5, and agar 20.
2.6. Time course study for biotransformation of all substrates
In time course experiment, the fine powder of each substrate
(4 g/L) was added into the medium after C. lini ST-1 was cultivated
for 24 h. Transformation reaction was carried out on a rotary shaker
(220 rpm) at 28 ◦C for 72 h. Time course for biotransformation of
all substrates were analyzed according to a method reported by
Koshimura with slight modification [17]. The samples (1 mL) were
withdrawn aseptically at every 6 h, extracted with ethyl acetate and
the level of transformation was monitored by HPLC.
2.3. Conditions of cultivation and transformation
Fungi were inoculated aseptically into 500 mL flasks with
100 mL of fresh fermentation medium containing the following
components (g/L): glucose 30, yeast extract 15, corn steep liquor 3,
KCl 1, (NH4)2HPO4 1, K2HPO4 1, and MgSO4·7H2O 1. The medium
was adjusted to pH 7.0 with 0.5 M NaOH or HCl solution. After cul-
tivated for 72 h at 30 ◦C on a rotary shaker, aliquots (8 mL) from
the seed flask were transferred aseptically to fresh fermentation
medium and cultivated for a further 24 h under the same conditions
as above. Then, the fine powder of substrate (4 g/L) was added to the
culture under sterile conditions and incubated for another 60–72 h.
2.7. Confirmation of the proposed dihydroxylation pathway
To further confirm the proposed dihydroxylation pathway,
biotransformations of intermediate products purified from trans-
formation extract with C. lini ST-1 were performed. After C. lini ST-1
was cultivated in fermentation medium for 24 h under the same
conditions as described in Section 2.3, the fine powder of puri-
fied products (2 g/L) was added into the medium. Transformation
reaction was carried out in 250 mL flasks containing 30 mL of fer-
mentation medium on a rotary shaker (220 rpm) at 28 ◦C for 72 h.
Samples were withdrawn at regular intervals and transformation
products were subsequently extracted with ethyl acetate for three
times and dried in vacuum. The analysis of product concentrations
were the same as described in Section 2.6.
2.4. Preparative fermentation
The same transformation was performed in three 500 mL flasks.
Each flask contained 100 mL of the fermentation medium. The
cultures were cultivated under the same conditions and then an
average of 1200 mg substrate was added to the three grown cul-
tures under sterile conditions. Following 60–72 h transformation,
the fungal mycelium and fermentation broth were extracted sep-
arately with ethyl acetate (3× 100 mL), dried with MgSO4 and

Y. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 120 (2015) 111–118
113
Table 1
1H NMR data for steroidal starting material (except compounds 1 and 2) and transformation products determined in DMSO-d6.
Reference data
Other significant signals
4-H
18-H
19-H
Androst-4-en-3,17-dione (3)
Testosterone (4)
5.64
5.61
0.84
0.68
1.17
1.14
4.43 (1H, d, J = 4.8 Hz, 17-OH);
3.44 (1H, td, J = 8.6, 4.8 Hz, 17-H)
Transformation products
15␣-Hydroxyandrost-4-en-3,17-dione (8)
5.63
5.61
0.86
0.87
1.18
1.27
4.94 (1H, d, J = 6.4 Hz, 15-OH);
4.19 (1H, m, 15-H);
2.90 (1H, dd, J = 19.1, 8.0 Hz, 16-H)
4.90 (1H, d, J = 6.5 Hz, 15-OH);
4.44 (1H, d, J = 6.7 Hz, 11-OH);
4.16 (1H, ddt, J = 10.0, 8.1, 6.5 Hz, 15-H);
3.81 (1H, m, 11-H);
11␣,15␣-Dihydroxyandrost-4-en-3,17-dione (9)
2.94 (1H, dd, J = 19.1, 8.0 Hz, 16-H)
4.49 (1H, d, J = 4.8 Hz, 17-OH);
4.42 (1H, d, J = 6.3 Hz, 15-OH);
3.82 (1H, tdd, J = 9.6, 6.2, 3.5 Hz, 15-H)
15␣-Hydroxytestosterone (10)
5.61
0.67
1.15
2.8. Identification of metabolites
Transformation of androst-4-en-3,17-dione (3) afforded two
products: 15␣-hydroxyandrost-4-en-3,17-dione (8) and 11␣,15␣-
dihydroxyandrost-4-en-3,17-dione (9) (Fig. 2). The 1H and 13C NMR
spectra of (8) were similar to those of (3). In 1H NMR, an additional
methine proton was observed as a result of the secondary alco-
hol effect at 4.19 (1H, m). This proton was correlated with H-14
and H-16 in the COSY spectrum. Therefore, ıH 4.19 was assigned
as H-15. Simultaneously, C-15 (ıC 68.55 ppm) was identified by
HSQC. This assignment was fully supported by downfield shifts
for C-14 (5.55 ppm) and C-16 (10.38 ppm). Additionally, the struc-
ture of (8) was confirmed by IR spectra which showed a strong
absorption peak of a hydroxyl group at 3413 cm−1. This finding was
further verified by accurate mass measurement (HRMS ESI: calc.
for M + H 303.1960 amu C19H26O3 obsd. 303.1952 amu). Moreover,
a NOESY correlation from H-15 to H-18 confirmed the configu-
ration of 15-OH was ␣. Unlike the 1H NMR spectrum of (8), that
of 11␣,15␣-dihydroxyandrost-4-en-3,17-dione (9) revealed a new
signal at 3.81 (1H, m) and significant downfield shifts for 19-methyl
(0.09 ppm) protons. This phenomenon confirmed the occurrence
of hydroxylation at C-11. This finding was fully supported by the
13C NMR spectrum of (9), which demonstrated downfield shifts
for C-9 (5.38 ppm) and C-12 (10.63 ppm). The COSY cross peak of
ıH 3.81 and H-9 also verified hydroxylation position. The NOESY
Characteristic patterns [23,24] and chemical shift values [25]
in the 1H and 13C NMR spectra in comparison to the starting
compounds were used to determine metabolite structure in com-
bination with DEPT and 2D NMR analysis to identify the nature
of the carbon (Tables 1–3). Spectra were recorded on a Bruker
AVANCE III 400 MHz Spectrometer, and all compounds were ana-
lyzed in DMSO-d6 using tetramethylsilane as the internal standard.
High resolution mass measurement (HRMS) was determined in
electrospray ionization (ESI) mode using a Platform ZMD 4000
spectrometer. Infra-red absorption spectra (Table 4) were recorded
directly on a Nicolet 380 FT-IR. The yields of all hydroxylation prod-
ucts converted from the tested steroidal substrates and separated
by column chromatography were showed in Table 5.
3. Results
3.1. Products of biotransformation and structural identification
With substrates of estradiol (1) and estrone (2), there is no
noticeable transformation by C. lini ST-1 for 120 h, and no consump-
tion of the starting substrates.
Table 2
1H NMR data for steroidal starting material (except compounds 1 and 2) and transformation products determined in DMSO-d6.
Reference data
Other significant signals
4-H
18-H
19-H
21-H
Canrenone (5)
16␣,17␣-Epoxyprogesterone (6)
Progesterone (7)
5.62
5.62
5.63
0.92
1.00
0.58
1.08
1.15
1.14
6.18 (2H, m, 6-H & 7-H)
3.93 (1H, s, 16-H)
1.97
2.06
Transformation products
15␣-Hydroxycanrenone (11)
5.60
5.59
0.91
0.92
1.08
1.18
5.02 (1H, d, J = 6.2 Hz, 15-OH);
4.11 (1H, tdd, J = 9.5, 6.2, 3.4 Hz, 15-H);
6.16 (1H, dd, J = 10.0, 2.7 Hz, 6-H);
6.68 (1H, dd, J = 9.7, 2.0 Hz, 7-H)
5.04 (1H, d, J = 6.4 Hz, 15-OH);
4.43 (1H, d, J = 6.5 Hz, 11-OH);
4.07 (1H, m, 15-H);
11␣,15␣-Dihydroxy-canrenone (12)
3.87 (1H, m, 11-H);
6.16 (1H, dd, J = 9.9, 2.6 Hz, 6-H);
6.66 (1H, dd, J = 9.8, 1.7 Hz, 7-H);
4.97 (1H, d, J = 7.1 Hz, 15-OH);
3.84 (2H, m, 15-H and 16-H)
4.99 (1H, d, J = 7.2 Hz, 15-OH);
4.36 (1H, d, J = 6.8 Hz, 11-OH);
3.88 (1H, d, J = 1.1 Hz, 16-H);
3.81 (2H, m, 11-H and 15-H)
4.58 (1H, d, J = 6.1 Hz, 15-OH);
3.81 (1H, m, 15-H)
15␣-Hydroxy-16␣,17␣-epoxyprogesterone (13)
5.61
5.60
1.05
1.02
1.16
1.25
1.97
1.97
11␣,15␣-Dihydroxy-16␣,17␣-epoxyprogesterone (14)
15␣-Hydroxyprogesterone (15)
5.62
0.57
1.14
2.05

114
Y. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 120 (2015) 111–118
Table 3
13C NMR data for transformation products determined in DMSO-d6.
Carbon atom
Compound
(8)
(9)
36.95
(10)
35.13
(11)
33.39
(12)
33.86
(13)
34.98
(14)
36.81
(15)
34.81
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
35.08
33.60
197.94
122.93
170.88
31.97
31.12
34.81
53.27
38.27
19.78
31.73
49.52
56.53
68.55
45.95
216.35
14.77
17.00
33.85
198.64
123.28
171.68
32.87
31.49
34.36
58.65
39.71
66.82
42.36
49.88
56.08
68.12
46.18
215.40
15.75
18.11
33.57
197.89
122.80
171.25
32.11
32.04
35.01
53.69
38.24
20.16
36.62
43.49
57.51
70.82
42.49
77.28
12.60
17.01
33.60
198.09
122.76
163.10
126.68
141.64
37.03
49.49
35.51
19.58
30.99
46.37
53.60
69.70
46.75
92.80
15.66
16.05
30.62
28.76
176.25
33.50
198.47
123.26
163.44
126.88
140.78
35.66
54.96
37.24
66.13
42.92
46.33
53.27
69.31
46.83
92.39
16.72
16.83
30.45
28.66
176.14
33.57
197.94
122.93
170.94
31.93
31.30
32.82
53.55
38.30
19.94
31.26
40.67
49.15
71.36
62.85
67.96
16.53
16.84
204.21
25.64
33.85
198.62
123.30
171.65
32.82
31.21
32.38
59.04
39.74
66.59
42.79
40.79
48.73
70.88
63.26
67.72
17.51
18.12
204.10
25.62
33.56
198.15
122.84
171.30
32.04
31.78
34.75
53.36
38.19
20.42
38.15
43.51
61.49
71.53
35.10
59.99
14.22
16.96
208.38
31.20
Table 4
Significant products absorption signals (cm−1).
Compound
OH
C
O
Lactone
15␣-Hydroxy-androst-4-en-3,17-dione (8)
11␣,15␣-Dihydroxy-androst-4-en-3,17-dione (9)
15␣-Hydroxytestosterone (10)
3413 br, 3510
3433 br
3412
1670, 1730
1632, 1740
1658
–
–
–
15␣-Hydroxycanrenone (11)
11␣,15␣-Dihydroxy-canrenone (12)
15␣-Hydroxy-16␣,17␣-epoxyprogesterone (13)
11␣,15␣-Dihydroxy-16␣,17␣-epoxyprogesterone (14)
15␣-Hydroxyprogesterone (15)
3427 br
3421 br
3450 br, 3521
3419
1647
1649
1659, 1699
1659, 1705
1660, 1693
1770
1767
–
–
–
3489
enhancement between 11-H and 18-H indicated the configuration
of 11-OH was ␣. The presence of dihydroxyl was confirmed by accu-
rate mass measurement (HRMS ESI: calc. for M + H 319.1909 amu
C19H26O4 obsd. 319.1912 amu).
Transformation of testosterone (4) resulted in the isolation
of one product (10) (Fig. 2). The structure of 15␣-hydroxy-
testosterone (10) was determined by comparing its spectra with
those of (4). The 1H NMR spectrum of (10) revealed two new
signals at 4.42 ppm (1H, d, J = 6.3 Hz) and 3.82 ppm (1H, tdd,
J = 9.6, 6.2, 3.5 Hz). This observation strongly suggested that a new
hydroxyl group was created. The proposed structure was also
supported by accurate mass measurement (HRMS ESI: calc. for
M + H 305.2117 amu C19H28O3 obsd. 305.2104 amu), which showed
the mass of (10) was 16 large from that of (4). In 13C NMR and
DEPT spectra, a new signal (70.82 ppm) was appeared and indi-
cated methine directly attached to the hydroxyl group. Moreover,
the signal of C-15 (23.00 ppm) in 13C NMR spectra of (4) disap-
peared in that of (10), which suggested the presence of OH-15.
This was further supported by ␤-carbon downfield shifts for C-16
(12.59 ppm) and C-14 (7.36 ppm), and by ␥-carbon upfield shifts
for C-17 (2.57 ppm) and C-13 (1.09 ppm). Heteronuclear 2D NMR
(HSQC and HMBC) also supported the assignment of (10). Orien-
tation of OH-15 was decided by NOESY, which showed a cross
peak of H-15 and H-18. Since H-18 was known to be ␤ configura-
tion, H-15 must be ␤ configuration and OH-15 was ␣ configuration
consequently.
Conversion of canrenone (5) gave two hydroxyl compounds,
15␣-hydroxy-canrenone (11) and 11␣,15␣-dihydroxy-canrenone
(12) (Fig. 3). In the 1H NMR of (11), an additional methine pro-
ton at 4.11 ppm (1H, tdd, J = 9.5, 6.2, 3.4 Hz) was observed, which
was in the region specific for secondary alcohols. The proton signal
observed at 5.02 ppm (1H, d, J = 6.2 Hz) was assigned as a hydroxyl
group based on the COSY cross peak between ıH 5.02 ppm and
ıH 4.11 ppm. Moreover, cross peak between ıH 4.11 ppm and H-
14 in COSY confirmed that the hydroxyl group linked to C-15.
In the IR spectrum, the strong band of hydroxyl group visible
at 3427 cm−1 also indicated the presence of a hydroxyl group.
The OH stereochemistry was determined as ␣ configuration by
NOESY cross peak between H-15 and H-18. The occurrence of
Table 5
Products yields following chromatography.
Starting material
Transformation products
Yield (%)
Androst-4-en-3,17-dione (3)
4
5
15␣-Hydroxyandrost-4-en-3,17-
dione (8)
11␣,15␣-Dihydroxyandrost-4-
en-3,17-dione (9)
64
Testosterone (4)
Canrenone (5)
5
60
5
15␣-Hydroxytestosterone (10)
15␣-Hydroxycanrenone (11)
11␣,15␣-Dihydroxy-
canrenone (12)
22
47
16␣,17␣-Epoxyprogesterone
(6)
5
28
48
15␣-Hydroxy-16␣,17␣-
epoxyprogesterone (13)
11␣,15␣-Dihydroxy-16␣,17␣-
epoxyprogesterone (14)
Progesterone (7)
24
27
15␣-Hydroxyprogesterone (15)

Y. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 120 (2015) 111–118
115
Fig. 2. Transformation of estradiol (1), estrone (2), androst-4-en-3,17-dione (3) and testosterone (4) by C. lini ST-1.
monohydroxylation was further confirmed by accurate mass mea-
surement (HRMS ESI: calc. for M + H 357.2066 amu C22H28O4 obsd.
357.2067 amu). 11␣,15␣-Dihydroxy-canrenone (12) was identified
by comparing its spectra with those of compound (11). A new signal
detected in 1H NMR spectrum of the product indicated hydrox-
ylation was occurred at a methylene. This indication was further
confirmed by accurate mass determination (HRMS ESI: calc. for
M + H 373.2015 amu C22H28O5 obsd. 373.2018 amu). The 1H NMR
spectrum of (12) revealed that the 19-methyl signal underwent
a downfield shift (0.10 ppm), resulting in a resonance signal at
1.18 ppm which was signatory for C-11 hydroxylation. This assign-
ment was fully supported by downfield shifts for C-9 (5.47 ppm)
and C-12 (11.93 ppm). NOESY enhancement between H-11 and H-
18 indicated the configuration of 11-OH was ␣.
Incubation of 16␣,17␣-epoxyprogesterone (6) resulted in the
isolation of two steroidal metabolites (Fig. 3). The structure of
15␣-hydroxy-16␣,17␣-epoxyprogesterone (13) was identified by
comparing of its 1H and 13C NMR spectra with that of the starting
material. The 1H NMR spectrum of (13) demonstrated the integral
value of resonance signal at 3.84 ppm increased from 1 to 2, which
suggested the occurrence of monohydroxylation. This observation
was also confirmed by accurate mass spectrometry (HRMS ESI: calc.
for M + H 345.2066 amu C21H28O4 obsd. 345.2075 amu). The COSY
cross peak between ıH 3.84 ppm and H-14 revealed the position of
monohydroxylation on C-15. The structure was confirmed by the IR
spectrum, and a hydroxyl group was visible at 3450 and 3521 cm−1
.
NOESY correlation from H-15 to H-18 confirmed ␣ configuration of
OH-15. 11␣,15␣-Dihydroxy-16␣,17␣-epoxyprogesterone (14) was
Fig. 3. Transformation of canrenone (5), 16␣,17␣-epoxyprogesterone (6), progesterone (7) by C. lini ST-1.

116
Y. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 120 (2015) 111–118
Fig. 4. Time course of biotransformation of (3, 4, 5, 6 and 7) by C. lini ST-1.
verified by comparing of its 1H and 13C NMR spectra with those
of (13). The 1H NMR spectrum of (14) demonstrated a downfield
shift for the 19-methyl group (0.09 ppm) and the presence of a new
signal at 3.81 ppm. This phenomenon suggested that a hydroxyl-
ation might occur at position C-11. Accurate mass measurement
fully supported the structure of (14) (HRMS ESI: calc. for M + H
361.2015 amu C21H28O5 obsd. 361.2000 amu). The 13C NMR spec-
trum of (14) demonstrated ␤-carbon was shifted downfield for C-9
(5.49 ppm) and C-12 (11.53 ppm). This occurrence was consistent
with the substitution at C-11. The ␣ configuration of OH-11 was
confirmed by NOESY cross peak between H-11 and H-18.
a high conversion rate of 82.46% (Fig. 4A). While incubated pro-
gesterone (7) with C. lini ST-1 for the same time as that of (4),
the 15␣-hydroxy-progesterone (15) conversion rate of 39.64% was
obtained (Fig. 4B).
Hydroxylation of androst-4-en-3,17-dione (3), canrenone (5)
and 16␣,17␣-epoxyprogesterone (6) underwent a similar route of
dihydroxylation with C15 and C11 (Fig. 4C–E). In the first 12 h, the
rate of converting (3) to 15␣-hydroxyandrost-4-en-3,17-dione (8)
increased significantly to 66.96%. Meanwhile the rate of convert-
ing (8) to 11␣,15␣-hydroxyandrost-4-en-3,17-dione (9) increased
slightly, and only a small amount of (9) was generated. After
another 36 h of transformation, a majority of (8) was further con-
verted into (9). Finally, the conversion rate of (8) decreased to
5.87%, and the conversion rate of (9) increased to 75.64%. These
results demonstrated the formation of (9) from (3) was achieved
by first introducing hydroxyl groups into the 15␣-position, fol-
lowed by hydroxylation at the 11␣-position of the steroid
molecule. Using canrenone (5) and 16␣,17␣-epoxyprogesterone
(6) as substrates, generated a similar transformation pathway, 15␣-
hydroxyl products and 11␣,15␣-dihydroxyl products were formed
as well.
The mass of molecular ion of 15␣-hydroxy-progesterone
(15) (HRMS ESI: calc. for M + H 331.2273 amu C21H30O3 obsd.
331.2271 amu) was 16 large from that of (7), which indicated
hydroxylation occurred at the substrate (Fig. 3). IR signal at
3489 cm−1 also supported the occurrence of hydroxylation. The 1
H
NMR spectrum showed a methine proton was generated as a result
of the secondary alcohol present at ıH 3.81 ppm (1H, m), which
correlated with H-14 and H-16 in the COSY spectrum. The ste-
reochemistry of the hydroxyl group was established using NOESY
correlations. NOESY enhancement between H-18 and H-15 indi-
cated that the hydroxyl group at position 15 was ␣ configuration.
3.3. Confirmation of the proposed biotransformation pathway
3.2. Time course study for androst-4-en-3,17-dione (3),
testosterone (4), canrenone (5), 16˛,17˛-epoxyprogesterone (6),
and progesterone (7)
To further determine the dihydroxylation order of the three sub-
strates (3, 5 and 6), purified intermediates (8 and 13) obtained
from (3) and (6) were taken as substrates, and the dynamics of
the accumulation of dihydroxyl products were studied. As shown
in Fig. 5, using 15␣-hydroxy-androst-4-en-3,17-dione (8) or 15␣-
hydroxy-16␣,17␣-epoxyprogesterone (13) as substrate, resulted in
the accumulation of 11␣,15␣-dihydroxy-androst-4-en-3,17-dione
The products formation and substrates consumption were
determined by HPLC, and the results were shown in Fig. 4.
Transformation of testosterone (4) for 48 h resulted in C15 monohy-
droxylation, which produced 15␣-hydroxy-testosterone (10) with
Fig. 5. Time course of biotransformation of purified intermediates (8 and 13) by C. lini ST-1.

Y. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 120 (2015) 111–118
117
(9) or 11␣,15␣-dihydroxy-16␣,17␣-epoxyprogesterone (14) with
extended transformation time. After incubated with C. lini ST-
1 for 48 h, the substrates of (8) and (13) could be converted
almost completely to (9) and (14). From another perspective, these
results indicated that 11␣,15␣-dihydroxy-steroids were formed
from steroidal compounds following the subsequent introduction
of hydroxyl groups to the 15␣-position and then to the 11␣-
position of the steroid molecule. Moreover, both of the previous
report [26] and our previous study (Fig. 5C, not published) on dihy-
droxylation of DHEA by C. lini were consistent with this proposed
biotransformation pathway.
[34,35], but only the monohydroxylation product 15␣-hydroxy-
progesterone (15) with 27% yield (Table 5) was separated after
cultivation with C. lini ST-1. The oxygen group in position C11
was regarded as essential for anti-inflammatory action [36]. Also
15␣-hydroxylations were important reactions in the production
of gestodene, which was a progestogen with oral contracep-
tive properties [37]. Comparison of the transformation results
between 3-keto-steroids of (3–7) (Figs. 2 and 3) and 3-hydroxy-
steroid (16) (Fig. 1) indicated that the structure of the A ring of
the steroid molecule significantly influenced the position of the
introduced hydroxyl group. The replacing of 3-keto-configuration
to 3-hydroxy-configuration might lead to the change of the hydrox-
ylation position.
4. Discussions
This study demonstrated the effects of stereostructure on
steroid hydroxylation sites by C. lini ST-1, which guided the fur-
ther application of this biocatalyst. Also, a series of novel steroid
compounds were generated during the investigation of substrate
spectrum, which might be of interest in academia and clinical
medicine in the future. To the best of our knowledge, this is the
first report on the 11␣,15␣-dihydroxylation of steroid substrates
(3, 5 and 6) by microorganism. The various reactions and excellent
conversion rates obtained with C. lini ST-1 indicated that this strain
is a promising fungus and offers a potential new route to generate
novel steroids.
The results of our research strongly proved that the substrates
structure had remarkable influence on steroids hydroxylation with
C. lini ST-1. Using estranes including estradiol (1) and estrone
(2) as the tested substrates, no monohydroxylation or dihydrox-
ylation occurred on the substrate molecule. Biotransformation of
androstanes (3, 4 and 16) and pregnanes (5, 6 and 7) with C. lini
ST-1 showed that these steroidal substrates could undergo hydrox-
ylation on rings B, C and D at positions that corresponded to the
three possible sites (C7, C11 and C15). Moreover, all hydroxyl
groups synthesized by biotransformation were ␣ configuration,
which indicated that C. lini ST-1 exhibited higher stereoselectivity.
This stereo-selectivity provided the hydroxyl groups synthesized
by biotransformation with a remarkable advantage over those
obtained by chemosynthesis. After incubated with C. lini ST-1 for
48–72 h, the conversion rate of substrates (3, 4, 5 and 6) could reach
70–85%. The transformation process for some valuable products
obtained in this study will be optimized in our subsequent work,
through the addition of substrate cosolvent, use of cell permeabi-
lizing agents, establishment of organic-aqueous biphasic media,
and immobilization of catalyst. After process optimization, the sub-
strate concentration and products conversion rate could be further
enhanced.
Incubation of androstanes (3 and 4) with C. lini ST-1 displayed
different catalytic characteristics. Using androst-4-en-3,17-dione
(3) as substrate, a 5% yield of 15␣-hydroxyandrost-4-en-3,17-dione
(8) and a 64% yield of 11␣,15␣-dihydroxyandrost-4-en-3,17-dione
(10) could be obtained (Fig. 4C, Table 5). While testosterone (4) cul-
tivated with C. lini ST-1 merely resulted in monohydroxylation at
C15␣ position with a high yield of 60%, a dihydroxylation reaction
was not observed after 72 h biotransformation (Fig. 4A, Table 5).
Monohydroxylation of testosterone at C11␣ or C15␣ had been dis-
covered in other filamentous fungi [27,28], however hydroxylation
at C11␣ was not observed in C. lini ST-1. According to previous
reports, allylic hydroxylation at 6␤-position for 4-ene molecules
was the predominant steroidal transformation [15,29]. However,
no allylic oxidation and no 6␤-hydroxylation were observed among
the 4-ene substrates (3 and 4) investigated in this study. This could
be attributed to the different specificities of monooxygenases in
various strains. Interesting, 15␣-hydroxylation of 4-ene steroids
was characteristic of many species of the Fusarium genus [30]. Dihy-
droxylation at C11 and C15 was unprecedented and it seemed to
be novel when compared to the wide range of fungi which could
perform steroid dihydroxylation [31–33].
Acknowledgements
This work is financially supported by National High Technology
Research and Development Program of China (863 Program) (No.
2011AA02A211). Moreover, we thank Beijing Ernst NMR Technol-
ogy Co., Ltd. (Beijing, China) for acquiring and processing the NMR
data in this research.
References
[1] S.B. Mahato, S. Garai, Steroids 62 (1997) 332–345.
[2] W.Y. Tong, X. Dong, Recent Pat. Biotechnol. 3 (2009) 141–153.
[3] A. Weber, M. Kennekke, U. Klages, K. Nickisch, R. Rhode, US Patent 5,472,854
(1995).
[4] S.M. Roberts, J.P. Eddolls, A.J. Willetts, A. Atkinson, J.P. Murphy, GB Patent
2,318,790 (1996).
[5] P. Fernandes, A. Cruz, B. Angelova, H.M. Pinheiro, J.M.S. Cabral, Enzyme
Microb. Technol. 32 (2003) 688–705.
[6] P.C. Peart, K.P. McCook, F.A. Russell, F.R. William, P.B. Reese, Steroids 76
(2011) 1317–1330.
[7] H. Li, H.M. Liu, W. Ge, L. Huang, L. Shan, Steroids 70 (2005) 970–973.
[8] A.S. Lamm, A.R.M. Chen, W.F. Reynolds, P.B. Reese, Steroids 72 (2007)
713–722.
[9] M.I. Choudhary, S. Erum, M. Atif, R. Malik, N.T. Khan, Atta-ur-Rahman,
Steroids 76 (2011) 1288–1296.
[10] M.V. Donova, O.V. Egorova, Appl. Microbiol. Biotechnol. 94 (2012) 1423–1447.
[11] E. Huszcza, J. Dmochowska-Gładysz, Phytochemistry 62 (2003) 155–158.
[12] R. Arnell, R. Johannisson, J. Lindholm, T. Fornstedt, B. Ersson, A. Ballagi, K.
Caldwell, Prep. Biochem. Biotechnol. 37 (2007) 309–321.
[13] M.I. Choudhary, M.Y. Mohammad, S.G. Musharraf, M. Parvez, A. Al-Aboudi,
Atta-ur-Rahman, Steroids 74 (2009) 1040–1044.
[14] H.J. Ruijssenaars, E.M.G.M. Sperling, P.H.G. Wiegerinck, F.T.L. Brands, J. Wery,
J.A.M. de Bonta, J. Biotechnol. 131 (2007) 205–208.
[15] A. Swizdor, T. Kolek, A. Panek, N. Milecka, Curr. Org. Chem. 16 (2012)
2551–2582.
[16] W. Lu, L. Du, M. Wang, J. Wen, B. Sun, Y. Guo, Biochem. Eng. J. 32 (2006)
233–238.
[17] M. Koshimura, T. Utsukihara, A. Hara, S. Mizobuchi, C.A. Horiuchi, M.
Kuniyoshi, J. Mol. Catal. B: Enzym. 67 (2010) 72–77.
[18] E.M. Ahmed, Braz. J. Microbiol. 38 (2007) 224–229.
[19] V.V. Kollerov, A.A. Shutov, V.V. Fokina, G.V. Sukhodol’skaya, S.A. Gulevskaya,
M.V. Donova, Appl. Biochem. Microbiol. 46 (2010) 198–205.
[20] L.H. Huang, J. Li, G. Xu, X.H. Zhang, Y.G. Wang, Y.L. Yin, H.M. Liu, Steroids 75
(2010) 1039–1046.
[21] A. Romano, D. Romano, E. Ragg, F. Costantino, R. Lenna, R. Gandolfi, F.
Molinari, Steroids 71 (2006) 429–434.
[22] Y. Wu, H. Li, Z.M. Lu, H. Li, Z.M. Rao, Y. Geng, J.S. Shi, Z.H. Xu, Steroids 84
(2014) 70–77.
Incubation of the pregnanes (5, 6) revealed both 11␣ and 15␣-
hydroxylase activity (Fig. 3). The hydroxylation products obtained
from canrenone (5) and epoxyprogesterone (6) after 48 h biotrans-
formation were separated by column chromatography, with total
product yields of 69% and 76% (Table 5), respectively. Further-
more, in all recorded cases, the activity of 11␣-hydroxylase was
observed after 12–24 h, which suggested that this enzyme might be
induced. Progesterone (7) was a commonly used steroid substrate
[23] J.W. Blunt, J.B. Stothers, Org. Magn. Reson. 9 (1977) 439–463.

118
Y. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 120 (2015) 111–118
´
[24] D.N. Kirk, H.C. Toms, C. Douglas, K.A. White, J. Chem. Soc., Perkin Trans. 2
(1990) 1567–1594.
[25] J.E. Bridgeman, P.C. Cherry, A.S. Clegg, J.M. Evans, E.R. Jones, A. Kasal, V.
Kumar, G.D. Meakins, Y. Morisawa, E.E. Richards, P.D. Woodgate, J. Chem. Soc.
C (1970) 250–257.
[30] T. Kołek, A. Swizdor, J. Steroid Biochem. Mol. Biol. 67 (1998) 63–69.
[31] H. Li, Z.Z. Fu, H. Li, X.M. Zhang, J.S. Shi, Z.H. Xu, J. Ind. Microbiol. Biotechnol. 41
(2014) 1497–1504.
[32] H.N. Bhatti, R.A. Khera, Steroids 77 (2012) 1267–1290.
[33] M.A. Faramarzi, M. Badiee, M.T. Yazdi, M. Amini, M. Torshabi, J. Mol. Catal. B:
Enzym. 50 (2008) 7–12.
[26] T.G. Lobastova, S.A. Gulevskaya, G.V. Sukhodolskaya, M.V. Donova, Appl.
Biochem. Microbiol. 45 (2009) 617–622.
[34] A. Swizdor, T. Kolek, A. Panek, A. Białon´ ska, BBA Mol. Cell Biol. Lipids 1811
[27] P.C. Peart, W.F. Reynolds, P.B. Reese, J. Mol. Catal. B: Enzym. 95 (2013)
70–81.
[28] A. Al-Aboudi, M.Y. Mohammad, S.G. Musharraf, M.I. Choudhary,
Atta-ur-Rahman, Nat. Prod. Res. 22 (2008) 1498–1509.
[29] A.C. Hunter, P.W. Mills, C. Dedi, H.T. Dodd, J. Steroid Biochem. Mol. Biol. 108
(2008) 155–163.
(2011) 253–262.
[35] A. Manosroi, S. Saowakhon, J. Manosroi, J. Steroid Biochem. Mol. Biol. 108
(2008) 132–136.
[36] L. Sedlaczek, Crit. Rev. Biotechnol. 7 (1988) 187–236.
[37] S. Mao, X. Hu, B. Hua, N. Wang, X. Liu, F. Lu, Biotechnol. Lett. 34 (2012)
2113–2117.