Highly regio- and stereoselective hydroxylation of vitamin D2 by CYP109E1

Article abstract of DOI:10.1016/j.bbrc.2020.01.091

Vitamin D2 is a form of vitamin D derived from mushrooms and plants which is structurally modified in the body due to the action of several enzymes. The resulting metabolites represent important compounds with potential bioactive properties. However, they are poorly studied and their availability is mostly limited. In order to identify new enzymes capable of producing vitamin D2 metabolites, we investigated a bacterial P450 monooxygenase, CYP109E1, which was previously shown to be a vitamin D3 hydroxylase. It was found that CYP109E1 catalyzes a vitamin D2 two-step hydroxylation at positions C24 and C25 resulting in the generation of 24(R),25-diOH VD2. Interestingly, the enzyme showed high selectivity towards vitamin D2, whereas it showed an unselective product pattern for the structurally similar vitamin D3. Our docking results for vitamin D2 and D3 revealed favorable hydroxylation positions for both substrates and suggested an explanation for the high selectivity of CYP109E1 towards vitamin D2. In addition, we established a whole-cell biocatalyst expressing CYP109E1 in Bacillus megaterium to produce 24(R),25-diOH VD2 and a production yield of 12.3 ± 1.2 mg/L was obtained after 48 h. To the best of our knowledge, this is the first report on the generation of 24(R),25-diOH VD2 by a microbial biocatalyst allowing a low-cost and eco-friendly production of this pharmaceutically interesting and expensive metabolite from the relatively cheap substrate, VD2.

Full text of DOI:10.1016/j.bbrc.2020.01.091

Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications
journal homepage: www.elsevier.com/locate/ybbrc
Highly regio- and stereoselective hydroxylation of vitamin D2 by
CYP109E1
a
a
b
c
a, *
Natalia Putkaradze , Lisa K o€ nig , Lars Kattner , Michael C. Hutter , Rita Bernhardt
a
Institute of Biochemistry, Saarland University, D-66123, Saarbruecken, Germany
Endotherm Life Science Molecules, D-66123, Saarbruecken, Germany
Center for Bioinformatics, Saarland University, D-66123, Saarbruecken, Germany
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Vitamin D2 is a form of vitamin D derived from mushrooms and plants which is structurally modified in
the body due to the action of several enzymes. The resulting metabolites represent important com-
pounds with potential bioactive properties. However, they are poorly studied and their availability is
mostly limited. In order to identify new enzymes capable of producing vitamin D2 metabolites, we
investigated a bacterial P450 monooxygenase, CYP109E1, which was previously shown to be a vitamin
D3 hydroxylase. It was found that CYP109E1 catalyzes a vitamin D2 two-step hydroxylation at positions
C24 and C25 resulting in the generation of 24(R),25-diOH VD2. Interestingly, the enzyme showed high
selectivity towards vitamin D2, whereas it showed an unselective product pattern for the structurally
similar vitamin D3. Our docking results for vitamin D2 and D3 revealed favorable hydroxylation positions
for both substrates and suggested an explanation for the high selectivity of CYP109E1 towards vitamin
D2. In addition, we established a whole-cell biocatalyst expressing CYP109E1 in Bacillus megaterium to
produce 24(R),25-diOH VD2 and a production yield of 12.3 ± 1.2 mg/L was obtained after 48 h. To the best
of our knowledge, this is the first report on the generation of 24(R),25-diOH VD2 by a microbial bio-
catalyst allowing a low-cost and eco-friendly production of this pharmaceutically interesting and
expensive metabolite from the relatively cheap substrate, VD2.
Received 22 December 2019
Accepted 15 January 2020
Available online xxx
Keywords:
Cytochrome P450
CYP109
Vitamin D
Hydroxylase
24,25-diOH VD2
©
2020 Elsevier Inc. All rights reserved.
1
. Introduction
availability of these compounds. Due to a very high clinical
importance of VD metabolites, many previous studies aimed to
chemically synthesize these compounds for functional studies as
well as to characterize the enzymes involved in VD metabolism in
humans (reviewed by Sakaki et al. [2] and Kattner and Volmer [4]).
It is known that several membrane-bound P450 isoforms (CYP2R1,
CYP27A1, CYP27B1, and CYP24A1) contribute to VD metabolism
carrying out the crucial oxidation steps, including 25-, 24- and 1-
hydroxylations [3]. Moreover, a few P450s from bacterial origin
have been identified to be capable of metabolizing VD and its de-
rivatives such as CYP105A1 from Streptomyces griseolus, CYP107
from Pseudonocardia autotrophica and CYP109E1 from Bacillus
megaterium [5e7]. These microbial enzymes are particularly
interesting as biocatalysts since they enable the production of VD
metabolites in an easy and environment-friendly manner, in
contrast to classical chemical synthesis. So far these bacterial P450
systems have been mostly investigated for their activity towards
VD3 and there is only very little known about VD2 metabolizing
P450s. Although VD2 is considered to be less potent than VD3, it is
metabolized in the human body and shows significant effects on
The term vitamin D (VD) refers to two structurally and func-
tionally related secosteroids, vitamin D2 (VD2, ergocalciferol) and
D3 (VD3, cholecalciferol). VD2 is a derivative of ergosterol in fungi
and yeast whereas VD3 is photosynthesized from 7-
dehydrocholesterol in the skin due to the action of solar ultravio-
let B radiation [1]. Both forms are hormone precursors lacking
biological activity on their own. They are activated and further
modified by cytochrome P450 monooxygenases (P450s) [2,3]. The
most potent form of VD, 1,25-diOH VD3, is known to be crucial for
maintaining calcium homeostasis in the body but can also regulate
cell proliferation and differentiation [2]. Apart from this, a high
number of VD metabolites have been isolated. However, their
physiological functions are mostly unknown due to poor
*
Corresponding author. Institute of Biochemistry, Saarland University, Campus, B
2.2, D-66123, Saarbruecken, Germany.
E-mail address: ritabern@mx.uni-saarland.de (R. Bernhardt).
https://doi.org/10.1016/j.bbrc.2020.01.091
006-291X/© 2020 Elsevier Inc. All rights reserved.
0
Please cite this article as: N. Putkaradze et al., Highly regio- and stereoselective hydroxylation of vitamin D2 by CYP109E1, Biochemical and
Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2020.01.091

2
N. Putkaradze et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
the overall VD status in mammals when taken via certain food and
food supplements [8e11]. Consequently, physiological VD2 me-
tabolites are important compounds for further functional studies
although such studies are scarce due to the lack of efficient pro-
duction systems for these metabolites.
medium (24 g/L yeast extract, 12 g/L soytone, 0.5% glycerol (v/v),
12.5 g/L K HPO , 2.3 g/L KH PO , 10 g/mL tetracycline) with 1% of
2
4
2
4
m
the overnight culture (v/v) and were grown under shaking
ꢀ
(160 rpm) at 37 C until the OD578nm of 0.5 was reached. Then, the
expression of CYP109E1 was induced by addition of 5 g/L xylose and
ꢀ
In this study, we aimed to investigate the activity of the bacterial
P450, CYP109E1, towards VD2 in order to explore a new biocatalyst
capable of producing VD2 metabolites. CYP109E1 has been exten-
sively studied by our group and was identified as versatile mono-
oxygenase acting on chemically different compounds, among
others on cholesterol and VD3 [7,12,13]. The wild-type CYP109E1
showed, however, an unselective product pattern for VD3 hy-
droxylation [7]. Owing to the fact that VD2 differs from VD3
exclusively at the hydrocarbon side chain, we supposed that pro-
nounced differences might occur in the metabolism of VD2 in
comparison to VD3 and examined the activity of CYP109E1 for VD2
hydroxylation. Our experiments revealed that CYP109E1 acts as a
highly regio- and stereoselective VD2 hydroxylase and is able to
synthesize one of the physiological VD2 metabolites, 24(R),25-diOH
VD2, via two-step hydroxylation. A CYP109E1 expressing system in
B. megaterium was established that was able to produce 24(R),25-
diOH VD2 in an inexpensive and sustainable way.
the culture was further incubated at 30 C and 150 rpm. After 22 h
of incubation, the cells were harvested by centrifugation at 4500 x g
for 15 min followed by a washing step with 50 mM potassium
phosphate buffer (pH 7.4) containing 2% glycerol (v/v). Whole-cell
conversion of 160 mM VD2 (dissolved in 45% 2-hydroxypropyl-b-
cyclodextrin and 4% Quillaja saponin solution) was carried out in a
25 mL reaction volume (90 ± 10 g wet cells in 1 L potassium
ꢀ
phosphate buffer) in a rotary shaker at 30 C and 150 rpm. Samples
were taken at different time points, extracted with a triple volume
ꢀ
of ethyl acetate, dried under vacuum and stored at ꢁ 20 C until
HPLC analysis. For the production of the final product in mg
quantities for NMR analysis, the culture volume was upscaled to 1 L
and extracted with ethyl acetate.
2.5. RP-HPLC analysis
For the reversed-phase high-performance liquid chromatog-
raphy (HPLC) analysis the samples were dissolved in 200
acetonitrile. The measurements were carried out on a Jasco system
Pu-980 HPLC pump, AS-950 sampler, UV-975 UV/Vis detector, LG-
mL
2
. Materials and methods
(
2
.1. Strains and chemicals
980-02 gradient unit (Gross-Umstadt, Germany)) equipped with an
EC 125/4 NUCLEODUR 100-5 C18 column from Macherey Nagel
(Dueren, Germany). The column temperature was adjusted to 40 C
ꢀ
For purification, CYP109E1 was expressed in Escherichia coli C43
(
DE3) cells (Lucigen, Middleton, WI, USA) harboring the plasmid
and the flow rate was set at 1 mL/min. A linear gradient of 65.8%
aqueous acetonitrile to 100% pure acetonitrile for 19 min followed
by 100% acetonitrile for 12 min was used. The UV detection of VD2
and its metabolites was accomplished at 265 nm. The production of
25-OH VD2 and 24(R),25-diOH VD2 was quantified using the peak
areas (mV) of the metabolites on HPLC chromatograms and the
respective calibration curves.
pET17b.CYP109E1 [7]. E. coli strains JM109 and BL21 were used for
the expression of bovine redox proteins. Whole-cell conversions of
VD2 were performed using B. megaterium MS941 strain harboring
the plasmid pSMF2.1.CYP109E1. VD2, 25-OH VD2, 24(R),25-diOH
VD2, 24(S),25-diOH VD2 were provided by Endotherm (Saar-
bruecken, Germany). All other reagents and chemicals were from
standard sources and of highest purity available.
2.6. Isolation of VD2 metabolite and NMR characterization
2.2. Protein expression and purification
For the HPLC separation of the final product P2 from the cell
The heterologous expression and purification of CYP109E1 was
extract, the samples (10 mg) were dissolved in 0.5 mL chlor-
oform:methanol mixture (99:1). The separation measurements
were carried out on a Varian system (PrepStar SD-1, 320 UV/Vis
detektor Model) equipped with a Dynamax-60 Å preparative col-
umn (Si-83-121-C) from Rainin Instruments (MA, USA). The column
temperature was at RT and the flow rate was set at 17 mL/min. The
UV detection of VD2 and its metabolites was accomplished at
performed as described previously [12]. The bovine adrenodoxin
reductase (AdR) and the truncated bovine adrenodoxin (Adx4e108
)
were expressed and purified as reported by Sagara et al. [14] and
Uhlmann et al. [15].
2
.3. In vitro conversion
2
6
54 nm. NMR spectra were recorded in D -Acetone with a Bruker
1
13
The in vitro turnover of VD2 was performed with a reconstituted
system, containing CYP109E1 (1 mM), bovine AdR and Adx4-108 in a
DRX 500 spectrometer. The structure of P2 was assigned by H/
NMR.
C
molar ratio of 1:2:20. For the continuous electron supply, a NADPH
regeneration system was used consisting of 1 U glucose-6-
phosphate-dehydrogenase, 5 mM glucose-6-phosphate and 1 mM
2.6.1. P2 (24(R),25-diOH VD2):
1
6
H NMR (D -Acetone, 500 MHz): d 0.59 (s, 3xH-18), 1.04 (d,
MgCl
trin/water or 25-OH VD2, dissolved in ethanol, were added to the
reaction mixture with a final concentration of 80 M. The reactions
2
. Either VD2, dissolved in 45% 2-hydroxypropyl-
b
-cyclodex-
J ¼ 6.5 Hz, 3xH-21), 1.14 (d, J ¼ 2.5 Hz, 3xH-26 þ 3xH-27), 1.23 (s,
3xH-28), 1.29e2.24 (m, 16H), 2.41 (m, H), 2.54 (m, H), 2.86 (m, H),
3.19 (s, OH), 3.38 (s, OH), 3.69 (d, J ¼ 4.5 Hz, OH), 3.79 (m, H-3), 4.74
(brs, H-19), 5.03 (brs, H-19), 5.55 (dd, J ¼ 15.5 and 8.5 Hz, H-22),
5.69 (d, J ¼ 15.5 Hz, H-23), 6.06 (d, J ¼ 11.0 Hz, H-6), 6.23 (d,
J ¼ 11.0 Hz, H-7).
m
were carried out in 250
tassium phosphate buffer (with 10% glycerol, pH 7.4) at 30 C and
m
L total reaction volume with 50 mM po-
ꢀ
7
50 rpm. The reactions were quenched after 6 and 90 min and
13
extracted with ethyl acetate.
C NMR (D
C-21), 22.91, 23.26, 24.16, 25.36, 25.50, 28.52, 33.19 (CH
36.60 (CH , C-2), 41.07 (CH, C-12), 41.20 (CH , C-20), 46.41 (C, C-13),
7.13 (CH , C-4), 57.04 (CH, C-17), 57.28 (CH, C-14), 69.61 (CH, C-3),
74.71 (C, C-25), 77.00 (C, C-24), 112.19 (CH , C-19), 118.77 (CH, C-7),
6
-Acetone, 125 MHz):
d
12.58 (CH
3
, C-18), 21.23 (CH
3
,
2
, C-1),
2.4. Whole-cell conversion
2
2
4
2
ꢀ
B. megaterium cells were grown over night at 37 C in 30 mL LB
medium supplemented with 10 g/mL tetracycline. The main cul-
tures were prepared by inoculating of 50 mL of an enriched
2
m
122.15 (CH, C-6), 125.76 (CH, C-22), 133.32 (C, C-5), 135.41 (CH, C-
23), 137.48, (C, C-8), 141.62 (C, C-10).
Please cite this article as: N. Putkaradze et al., Highly regio- and stereoselective hydroxylation of vitamin D2 by CYP109E1, Biochemical and
Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2020.01.091

N. Putkaradze et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
3
2
.7. Computational methods
structure than that of VD3 and cholesterol. Therefore, we have
focused on VD2, since, in contrast to VD3, it has a methyl group at
C24 and a double bond between C22 and C23 (Scheme 1). Position
24 in VD2 is highly reactive since it is both a tertiary carbon and an
allylic position.
The in vitro conversion of VD2 by a reconstituted CYP109E1
system over time showed that the substrate was at first converted
into product P1, which was the main product after 6 min of reac-
tion. P1 was then further metabolized into P2 (Fig. 1). Thus, the
enzyme was capable of converting VD2 into one main product, P2,
via an intermediate P1 which is in contrast to the conversion of
VD3, where 7 products are generated [7].
The crystal structure of CYP109E1 in complex with testosterone
(
PDB code 5L94) was chosen as receptor [16]. Chain A was prepared
using AutoDockTools (version 1.5.6r3) [17]. Protonation states of
the histidine residues were assigned manually by visual inspection
of the hydrogen-bonding network. Remaining hydrogen atoms and
Kollman charges were added by the program. Atomic charges for
the heme group were taken from previous docking studies [18]. The
molecular structures of the ligands were generated in the same way
as described for these studies. Further preparation steps for docking
were carried out with AutoDockTools. In order to comprise the
entire binding pocket, the rectangular grid box had an extension of
In order to produce mg amounts of P2 for its structural deter-
mination, we have further investigated the CYP109E1-mediated
conversion of VD2 in vivo using B. megaterium MS941. This strain
was applied in our previous studies to establish effective whole-cell
biocatalysts for the production of important steroidal compounds,
such as 15-hydroxycyproterone acetate, cortisone and 25-OH VD3
[7,23,24]. The CYP109E1 whole-cell biocatalyst was successfully
applied in this study for the conversion of VD2 and accomplished
the production of P2 via the intermediate P1, similar to the in vitro
transformation (Fig. 1B and C). Thus, the in vivo system was iden-
tified to be useful for the production of VD2 metabolites of
CYP109E1 and was subsequently used to generate sufficient
amounts of P2 for its structure determination. The P2 product was
purified from the cell extracts via preparative HPLC and was
5
2 ꢂ 50 ꢂ 72 point with a default grid spacing of 0.375 Å centered
above the heme. The actual docking was carried out by AutoDock
version 4.2), applying the Lamarckian genetic algorithm for a total
(
of 250 docking runs [17]. Otherwise default parameter settings
were used. Resulting conformations of VD2 and VD3 were inspec-
ted manually using VMD (version 1.9.1) [19].
3
. Results and discussion
The VD metabolism represents a challenging and exciting
research area for more than 40 years since VD is an important
prohormone showing high impact on human health. It is a fat-
soluble vitamin, occurs in form of VD2 and VD3 in nature and
plays a key role in maintaining calcium homeostasis [20,21]. The
VD forms undergo several enzymatic oxidations in the human
body and are transformed into different molecules, such as side-
chain modified VD derivatives. One of these compounds, 25-OH
VD3, is known to be the major circulating VD metabolite under
normal conditions, whereas several other metabolites, such as
1
13
analyzed by NMR. The H and C NMR identified P2 as dihy-
droxylated VD2 metabolite with hydroxyl groups at C24 and C25
(see NMR Data). To confirm the stereochemistry of the C24 hy-
droxyl, we compared P2 with authentic standards of 24(R),25-diOH
VD2 and 24(S),25-diOH via RP-HPLC. In addition, we compared the
intermediate product P1 with the authentic standard of 25-OH
VD2. The retention times of the peaks strongly suggested that P1
is 25-OH VD2 and the final product is 24(R),25-diOH VD2 (Fig. 2 A
and B, Scheme 2), which is one of the metabolites of VD2 detected
in animal plasma [10]. It is known, that 24(R),25-diOH VD2 can be
produced either from 24-OH VD2 or 25-OH VD2 whereby the latter
one being the favored precursor [25]. We have checked the catalytic
activity of CYP109E1 towards 25-OH VD2 and observed the for-
mation of 24(R),25-diOH VD2 in vitro (data not shown). Unfortu-
nately, 24-OH VD2 was not commercially available for an in vitro
activity assay with CYP109E1.
2
4,25-diOH VD2 and 25,26-diOH VD2, exist to a lesser extent in
the plasma [20,22]. Although the majority of the side-chain
modified metabolites are believed to be less potent than the
most active form, 1,25-diOH VD3, they occur in a considerable
amount but their physiological roles are not fully discovered yet
[
8,10]. One issue regarding their functional investigation could be
the scarce availability of VD metabolites, which can be overcome
by chemical or enzymatic synthesis of these compounds. Given
that chemical methods are challenging and harmful for the envi-
ronment, exploring biocatalytic alternatives for the production of
these physiological metabolites is highly desirable and of great
biotechnological importance. P450s are a superfamily of bio-
catalysts which are able to act on VD and its metabolites and to
produce complex derivatives under mild reaction conditions. Be-
sides the mammalian P450s, there are a small number of microbial
P450 enzymes that were characterized as VD hydroxylases [5e7].
One of those, CYP109E1, was intensively studied in our laboratory
and its crystal structure was solved in a substrate-free as well as in
substrate-bound form with the steroidal substrate, testosterone
To shed light on the experimentally observed selectivity of
CYP109E1 towards VD2 and differences to the conversion of VD3,
molecular docking calculations were performed using the crystal
structure of CYP109E1. The obtained conformations of VD2 suggest
that in the first step VD2 can be hydroxylated in both positions
(24(R) and 25), whereby position 24(R) is conformationally favored
(Fig. 3A). However, upon the formation of 24(R)-OH VD2, no further
reaction is expected, because no appropriate conformation that
would allow hydroxylation at position 25 was observed. On the
other hand, viewed from the point of the most frequently adapted
docking conformations, the subsequent hydroxylation of 25-OH
VD2 to 24(R),25-diOH VD2 is favored supporting the experimen-
tally observed results. In contrast, when looking at the docked
conformations for VD3, the most often adopted conformation of the
substrate, which was also energetically most favorable, showed the
positions 22 and 23 nearer to the heme iron than the corresponding
position 24 (Fig. 3B). It was further found that hydroxylation at all
three conceivable positions (24(R), 24(S), and 25) may occur
although these positions were energetically less favorable. This
fully supports the observed production of several metabolites of
VD3 by CYP109E1. Thus, the obtained docking conformations
explain the selectivity of the enzyme towards VD2 and VD3
hydroxylation.
[
16]. Besides being able to convert VD3 and testosterone, this
enzyme was found to efficiently act on cholesterol, another ste-
roidal compound with a hydrocarbon side chain [13]. Interest-
ingly, the enzyme showed low selectivity for VD3 converting this
compound into several (4 main and 3 minor) products [7],
whereas much higher selectivity was observed for cholesterol
conversion resulting in the generation of only two main products
[
13]. It was found that CYP109E1 prefers hydroxylation at the side-
chain positions 24(S) and 25 for both substrates. Such similarities
were not surprising since cholesterol is a precursor of VD3 and
they do not differ from each other in the structure of the side-
chain (Scheme 1). In this context, it was interesting to examine
the activity of CYP109E1 towards another steroidal compound
possessing a hydrocarbon side chain with another chemical
Please cite this article as: N. Putkaradze et al., Highly regio- and stereoselective hydroxylation of vitamin D2 by CYP109E1, Biochemical and
Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2020.01.091

4
N. Putkaradze et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
Scheme 1. Chemical structures of cholesterol (A), VD3 (B) and VD2 (C).
Fig. 1. Biotransformation of VD2 by CYP109E1: HPLC product pattern of the in vitro (A) and whole-cell conversion (B), and observed HPLC product peaks of P1 and P2 after 4, 24 and
8 h whole-cell conversion (C). The data represent mean values of three independent experiments.
4
Fig. 2. Comparison of HPLC retention times of: product P1 and the authentic standard of 25-OH VD2 (A), product P2 and the authentic standards of 24(R),25-diOH VD2 and
4(S),25-diOH VD2 (B) and production yield of VD2 metabolites by the CYP109E1-containing whole-cell system after 4, 24 and 48 h (C). The data represent mean values of three
independent experiments.
2
Scheme 2. Proposed reaction pathway of VD2 by CYP109E1.
Please cite this article as: N. Putkaradze et al., Highly regio- and stereoselective hydroxylation of vitamin D2 by CYP109E1, Biochemical and
Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2020.01.091

N. Putkaradze et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
5
Acknowledgements
The authors thank Birgit Heider-Lips for the purification of AdR
and Adx4-108 and Dr. Josef Zapp for the NMR measurement.
References
[
[
[
[
1] M.F. Holick, Vitamin D: importance in the prevention of cancers, type 1 dia-
betes, heart disease, and osteoporosis, Am. J. Clin. Nutr. 79 (2004) 362e371,
https://doi.org/10.1093/ajcn/79.3.362.
2] T. Sakaki, N. Kagawa, K. Yamamoto, K. Inouye, Metabolism of vitamin D3 by
cytochromes P450, Front. Biosci. J. Virtual Libr 10 (2005) 119e134, https://
doi.org/10.2741/1514.
3] I. Schuster, Cytochromes P450 are essential players in the vitamin D signaling
system, Biochim. Biophys. Acta 1814 (2011) 186e199, https://doi.org/
10.1016/j.bbapap.2010.06.022.
4] L. Kattner, D.A. Volmer, Synthesis of low abundant vitamin D metabolites and
assaying their distribution in human serum by liquid chromatography-
tandem mass spectrometry (LC-MS/MS) as a new tool for diagnosis and risk
prediction of vitamin D related diseases, in: S. Gowder (Ed.), A Critical Eval-
uation of Vitamin
D - Basic Overview, Intech publishers, Rijeka, 2017,
pp. 103e127.
[
[
5] N. Sawada, T. Sakaki, S. Yoneda, T. Kusudo, R. Shinkyo, M. Ohta, K. Inouye,
Conversion of vitamin D3 to 1alpha,25-dihydroxyvitamin D3 by Streptomyces
griseolus cytochrome P450SU-1, Biochem. Biophys. Res. Commun. 320 (2004)
156e164, https://doi.org/10.1016/j.bbrc.2004.05.140.
6] Y. Fujii, H. Kabumoto, K. Nishimura, T. Fujii, S. Yanai, K. Takeda, N. Tamura,
A. Arisawa, T. Tamura, Purification, characterization, and directed evolution
study of a vitamin D3 hydroxylase from Pseudonocardia autotrophica, Bio-
chem. Biophys. Res. Commun. 385 (2009) 170e175, https://doi.org/10.1016/
j.bbrc.2009.05.033.
[
7] A. Abdulmughni, I.K. J oꢀ ꢀz wik, N. Putkaradze, E. Brill, J. Zapp, A.-
M.W.H. Thunnissen, F. Hannemann, R. Bernhardt, Characterization of cyto-
chrome P450 CYP109E1 from Bacillus megaterium as a novel vitamin D3 hy-
droxylase, J. Biotechnol. 243 (2017) 38e47, https://doi.org/10.1016/
j.jbiotec.2016.12.023.
[
[
8] D. Hartwell, C. Hassager, C. Christiansen, Effect of vitamin D2 and vitamin D3
on the serum concentrations of 1,25(OH)2D2, and 1,25(OH)2D3 in normal
subjects, Eur. J. Endocrinol. 115 (1987) 378e384, https://doi.org/10.1530/
acta.0.1150378.
9] B.W. Hollis, W.B. Pittard, Evaluation of the total fetomaternal vitamin D re-
lationships at term: evidence for racial differences, J. Clin. Endocrinol. Metab.
Fig. 3. A) The obtained docking conformations showing the hydroxylation positions of
vitamin D2 in position 25 (orange) and position 24 (purple). The respective distance to
the heme iron (depicted as large sphere) are both 3.1 Å, respectively. Position 25 in
both conformers is shown as sphere. Hydrogen-bonds to backbone carbonyl oxygens
are indicated as dashed green lines. B) The most often adapted and best scored docking
position of vitamin D3 shows position 23 (denoted as sphere) closest to the heme iron
59 (1984) 652e657, https://doi.org/10.1210/jcem-59-4-652.
[
[
10] R.L. Horst, T.A. Reinhardt, S.G. Reddy, Vitamin D metabolism, in: D. Feldman,
J.W. Pike, F.H. Glorieux (Eds.), Vitamin D, Elsevier Academic Press, London,
2005, pp. 15e36.
11] L.A. Houghton, R. Vieth, The case against ergocalciferol (vitamin D2) as a
vitamin supplement, Am. J. Clin. Nutr. 84 (2006) 694e697, https://doi.org/
10.1093/ajcn/84.4.694.
(
(
3.0 Å), whereas the expected hydroxylation positions 24 and 25 are more distant
>4 Å). (For interpretation of the references to colour in this figure legend, the reader is
referred to the Web version of this article.)
[12] N. Putkaradze, M. Litzenburger, A. Abdulmughni, M. Milhim, E. Brill,
F. Hannemann, R. Bernhardt, CYP109E1 is a novel versatile statin and terpene
oxidase from Bacillus megaterium, Appl. Microbiol. Biotechnol. 101 (2017)
8
379e8393, https://doi.org/10.1007/s00253-017-8552-6.
Given that VD metabolites are important and high-priced
compounds, we have further investigated the CYP109E1-
mediated conversion in vivo using B. megaterium MS941. We
determined the production of 25-OH VD2 as well as 24(R),25-diOH
VD2 by this whole-cell biocatalyst within 48 h. As depicted in
Fig. 2C, comparable amounts of hydroxy- and dihydroxy-VD2 me-
tabolites were detected after 24 h of reaction, whereas a produc-
tivity of 12.3 ± 1.2 mg/L was achieved for 24(R)25-diOH VD2 after
[
[
13] N. Putkaradze, M. Litzenburger, M.C. Hutter, R. Bernhardt, CYP109E1 from
Bacillus megaterium acts as a 24- and 25-hydroxylase for cholesterol, Chem-
biochem 20 (2019) 655e658, https://doi.org/10.1002/cbic.201800595.
14] Y. Sagara, A. Wada, Y. Takata, M.R. Waterman, K. Sekimizu, T. Horiuchi, Direct
expression of adrenodoxin reductase in Escherichia coli and the functional
characterization, Biol. Pharm. Bull. 16 (1993) 627e630, https://doi.org/
10.1248/bpb.16.627.
[
[
[
[
15] H. Uhlmann, V. Beckert, D. Schwarz, R. Bernhardt, Expression of bovine
adrenodoxin in E. coli and site-directed mutagenesis of/2 FE-2S/cluster li-
gands, Biochem. Biophys. Res. Commun. 188 (1992) 1131e1138, https://
doi.org/10.1016/0006-291x(92)91349-u.
4
8 h. Taking into account that the established whole-cell system
16] I.K. J oꢀ ꢀz wik, F.M. Kiss, Ł. Gricman, A. Abdulmughni, E. Brill, J. Zapp, J. Pleiss,
can be further optimized for the production of 24(R)25-diOH VD2,
it is a valuable tool and starting point for a sustainable production
of this interesting metabolite.
R. Bernhardt, A.-M.W.H. Thunnissen, Structural basis of steroid binding and
oxidation by the cytochrome P450 CYP109E1 from Bacillus megaterium, FEBS J.
283 (2016) 4128e4148, https://doi.org/10.1111/febs.13911.
17] G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell,
A.J. Olson, Autodock 4 and AutoDockTools 4: automated docking with selec-
tive receptor flexibility, J. Comput. Chem. 30 (2009) 2785e2791, https://
doi.org/10.1002/jcc.21256.
18] N. Putkaradze, F.M. Kiss, D. Schmitz, J. Zapp, M.C. Hutter, R. Bernhardt,
Biotransformation of prednisone and dexamethasone by cytochrome P450
based systems e identification of new potential drug candidates, J. Biotechnol.
Funding
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
242 (2017) 101e110, https://doi.org/10.1016/j.jbiotec.2016.12.011.
[
[
19] W. Humphrey, A. Dalke, K. Schulten, VMD - visual molecular dynamics, J. Mol.
Graph. 14 (1996) 33e38, https://doi.org/10.1016/0263-7855(96)00018-5.
20] H.F. DeLuca, The vitamin D system in the regulation of calcium and phos-
phorus metabolism, Nutr. Rev. 37 (1979) 161e193, https://doi.org/10.1111/
j.1753-4887.1979.tb06660.x.
Declaration of competing interests
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
[
21] R.C. Tuckey, C.Y.S. Cheng, A.T. Slominski, The serum vitamin D metabolome:
what we know and what is still to discover, J. Steroid Biochem. Mol. Biol. 186
Please cite this article as: N. Putkaradze et al., Highly regio- and stereoselective hydroxylation of vitamin D2 by CYP109E1, Biochemical and
Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2020.01.091

6
N. Putkaradze et al. / Biochemical and Biophysical Research Communications xxx (xxxx) xxx
(
2019) 4e21, https://doi.org/10.1016/j.jsbmb.2018.09.003.
Factories 14 (2015) 28, https://doi.org/10.1186/s12934-015-0210-z.
[24] L. Koenig, P. Hartz, R. Bernhardt, F. Hannemann, High-yield C11-oxidation of
hydrocortisone by establishment of an efficient whole-cell system in Bacillus
megaterium, Metab. Eng. 55 (2019) 59e67, https://doi.org/10.1016/
j.ymben.2019.06.005.
[
[
22] R.L. Horst, E.T. Littledike, J.L. Riley, J.L. Napoli, Quantitation of vitamin D and
its metabolites and their plasma concentrations in five species of animals,
Anal. Biochem. 116 (1981) 189e203, https://doi.org/10.1016/0003-2697(81)
9
0344-4.
23] F.M. Kiss, M.T. Lundemo, J. Zapp, J.M. Woodley, R. Bernhardt, Process devel-
opment for the production of 15 -hydroxycyproterone acetate using Bacillus
megaterium expressing CYP106A2 as whole-cell biocatalyst, Microb. Cell
[25] N.J. Koszewski, T.A. Reinhardt, J.L. Napoli, D.C. Beitz, R.L. Horst, 24,26-
b
Dihydroxyvitamin D2:
a unique physiological metabolite of vitamin D2,
Biochemistry 27 (1988) 5785e5790, https://doi.org/10.1021/bi00415a058.
Please cite this article as: N. Putkaradze et al., Highly regio- and stereoselective hydroxylation of vitamin D2 by CYP109E1, Biochemical and
Biophysical Research Communications, https://doi.org/10.1016/j.bbrc.2020.01.091