Analysis of enantiofacial selective epoxidation catalyzed by flavin-containing monooxygenase Lsd18 involved in ionophore polyether lasalocid biosynthesis

Article abstract of DOI:10.1246/cl.140721

Enzymatic epoxidation represents a key biosynthetic transformation in the construction of polyether skeletons. A single flavin-containing monooxygenase, Lsd18, is involved in ionophore polyether lasalocid biosynthesis and participates in the enantioselective epoxidations of the diene precursor. Biotransformation studies utilizing structurally simplified monoolefin analogs with different substitution patterns revealed important structural requirements for the enantiofacial selectivity of Lsd18-catalyzed epoxidations. These results enabled us to propose a substrate binding model of Lsd18, which was applied to the biosynthesis of other polyethers.

Full text of DOI:10.1246/cl.140721

Received: July 30, 2014 | Accepted: August 4, 2014 | Web Released: August 8, 2014
CL-140721
Analysis of Enantiofacial Selective Epoxidation Catalyzed by Flavin-containing
Monooxygenase Lsd18 Involved in Ionophore Polyether Lasalocid Biosynthesis
Gaku Suzuki,1 Atsushi Minami,1 Mayu Shimaya,1 Takeshi Kodama,2 Yoshiki Morimoto,2 Hiroki Oguri,1 and Hideaki Oikawa*1
1Division of Chemistry, Graduate School of Science, Hokkaido University, Sapporo, Hokkaido 060-0810
2Department of Chemistry, Graduate School of Science, Osaka City University, Osaka 558-8585
(E-mail: hoik@sci.hokudai.ac.jp)
Enzymatic epoxidation represents a key biosynthetic trans-
formation in the construction of polyether skeletons. A single
flavin-containing monooxygenase, Lsd18, is involved in iono-
phore polyether lasalocid biosynthesis and participates in the
enantioselective epoxidations of the diene precursor. Biotransfor-
mation studies utilizing structurally simplified monoolefin analogs
with different substitution patterns revealed important structural
requirements for the enantiofacial selectivity of Lsd18-catalyzed
epoxidations. These results enabled us to propose a substrate
binding model of Lsd18, which was applied to the biosynthesis of
other polyethers.
dinucleotide and catalyzes two rounds of epoxidations on the
olefin moieties of diene precursor 2 to afford the corresponding
bisepoxide 3 (Scheme 1A). A homologous epoxidase is found in
other ionophore polyether biosynthetic gene clusters, suggesting
that a single epoxidase catalyzes multiple rounds of enantioselec-
tive epoxidations in biosyntheses (Scheme 1B).2,11 Intriguingly,
some epoxidases catalyze the epoxidation of polyene precursors in
a unique and enantioselective manner; for example, in monensin
biosynthesis, a single epoxidase, MonCI, installs (R,R)- and (S,S)-
epoxides on the internal and terminal olefin moieties, respectively.1
Enantioselective epoxidation is an intriguing topic in polyether
construction, because it defines the stereochemistry of polyethers.
However, difficulties regarding the preparation of functionally
active epoxidases and putative polyene precursors have been a
bottleneck in determining the functions of the epoxidases.
A previous study10 and Leadlay’s observations12 showed that
polyether formation occurs during chain elongation by polyketide
synthase (PKS). In lasalocid biosynthesis, the linear polyketide
chain bound to acyl carrier protein enter the active site of
epoxidase Lsd18 from the diene terminal. Recently, a biotrans-
formation system was established to investigate the function of
Lsd18 by utilizing an Lsd18-overexpressed host, and it was found
that simple monoolefin 6a with a sterically demanding moiety
such as a cyclopentyl group was accepted by Lsd18 as an
alternative substrate.4 The highly stereoselective epoxidation of
6a, which mimics the C19-C24 portion of 2, provides an
opportunity to investigate important structural motifs in the
enzymatic enantiofacial epoxidation of structurally simplified
analogs. In this study, several substrate analogs, 6b-6f, were
designed with different substitution patterns on the olefin moiety
Polyethers such as ionophore polyethers, ladder polyethers,
and oxasqualenoids constitute a structurally unique class of natural
products that contain the polycyclic polyether skeleton. The
structural diversity of the polyether skeleton is primarily dictated
by the number of ether rings and the manner in which the ring
closure occurs. The unique core structure has attracted many
chemists, and extensive biosynthetic studies concerning the poly-
ether skeleton have been reported. Following the initial proposal of
the unified biosynthetic model in 1983,1 Leadlay’s group identified
two key enzymes, epoxidase and epoxide hydrolase, responsible for
polyether formation.2 In 2008, an in vitro enzymatic transformation
was reported for the skeletal construction of lasalocid A (1)
catalyzed by epoxide hydrolase Las19.3 Additionally, the functions
of key enzymes, epoxidase Lsd184 and Lsd19,5-7 and a pair of
epoxide hydrolases MonBI/MonBII8,9 involved in the biosynthesis
of 1 and monensin A (4) (Scheme 1) were also elucidated.10 Lsd18
is a flavoprotein that contains tightly bound flavin adenine
(A) lasalocid biosynthesis
(R, R) (R, R)
Lsd18
Lsd19
O
O
18
22
24
OH
O
19
23
H O
epoxidation
cyclization
O
OH
H
O
O
OH
OH
OH
CO2H
H
OH
S
OH
diene precursor 2
bisepoxide 3
lasalocid A (1)
Enz
(B) monensin biosynthesis
Enz
S
O
OH OH OH
O
O
triene precursor 5
HO
H
(R, R) (R, R)
(S, S)
MonBI/
MonBII
O
O
O
H
O
O
HO
H
H
MonCI
O
O
O
OH
OMe
O
epoxidation
cyclization
O
O
triepoxide
monensin A (4)
HO
Scheme 1. Enzymatic polyether construction of (A) lasalocid and (B) monensin.
Chem. Lett. 2014, 43, 1779–1781 | doi:10.1246/cl.140721
© 2014 The Chemical Society of Japan | 1779

a: R1 = Me, R2 = Me, R3 = H
b: R1 = H, R2 = Me, R3 = Me
c: R1 = H, R2 = Me, R3 = H
d: R1 = H, R2 = H, R3 = Me
e: R1 = Me, R2 = H, R3 = H
f: R1 = H, R2 = H, R3 = H
Non-enzymatic
cyclization
R1
R1
R1
Lsd18
O
R2
R3
R2
R2
1
4
O
5
H
OH
R3
OH
OH
R3
OH
6a−6f (racemic)
7a−7d
8a−8d
R1
R1
O
R2
R3
R2
O
H
R3
OH
ent-7a−ent-7d
ent-8a−ent-8d
Scheme 2. Epoxidation of tested substrates with Lsd18-overexpressed Rhodococcus erythropolis.
Me
H
OH
Me
H
OH
H
H
Me
Me
OH
Me
Me
OH
O
O
O
O
O
O
H
H
H
H
H
H
Me
OH
H
H
Me
OH
H
H
H
H
cis/trans
cis/trans
cis/trans
cis/trans
cis/trans
cis/trans
8c-1: cis
8c-2: trans
ent-8c-1: cis
ent-8c-2 : trans
8d-1: cis
8d-2: trans
ent-8d-1: cis
ent-8d-2: trans
8b-1: cis
8b-2: trans
ent-8b-1: cis
ent-8b-2: trans
(i) Authentic 8c
(ii) Authentic 8d
(iii) Authentic 8b
cis
trans
cis*
trans*
cis*
trans*
8b-2
8c-1
ent-8c-2
8c-2
8b-1
8d-1 8d-2
ent-8d-2
ent-8c-1
ent-8b-1
ent-8b-2
(v) Lsd18 reaction with 6d
(iv) Lsd18 reaction with 6c
(vi) Lsd18 reaction with 6b
8c-2
/min
8d-1
8d-2
ent-8b-1
8b-1
8b-2
ent-8b-2
8c-1
ent-8d-2
ent-8c-1
ent-8c-2
#
130
140
150
72
74
76
190
200
210
/min
/min
Figure 1. GC-MS profiles of authentic samples (i) 8c, (ii) 8d, and (iii) 8b, and Lsd18-catalyzed reaction products with (iv) 6c, (v) 6d, and (vi) 6b.
*: exchangeable. #: impurity.
in order to examine the substrate specificity and enantiofacial
selectivity in Lsd18 epoxidation (Scheme 2).
Table 1. Summary of Lsd18-catalyzed epoxidation exchangeable
Relative
Conversion
/%
ee/% ee
Simple olefins 6b and 6c were synthesized from the corre-
sponding γ,δ-unsaturated ester (Johnson orthoester Claisen rear-
rangement product) according to a previous report4 (Scheme S1) or
from the ester with a terminal alkyne13 (Scheme S2). To obtain
authentic samples, epoxidation of 6b-6d with mCPBA was carried
out. The resulting epoxides were prone to epoxide-opening reac-
tions, thereby yielding cyclization products 8b-8d (Schemes S1
and S2). Hardly separable C6 diastereomers of 8c-1 and 8c-2
were isolated by HPLC following their conversion into the
corresponding (R)-2-methoxy-2-(1-naphthyl)propionate (MNPA)
esters, whose structures were determined by extensive NMR
analysis. Asymmetric epoxidation using Shi’s catalyst14 yielded the
enantiomerically enriched cyclization products. Using the synthetic
standards, we established the appropriate separation conditions
of three sets of stereoisomers 8b-8d by GC-MS using a chiral
capillary column. Chiral GC analysis of the asymmetric epoxida-
tion products revealed enantiomerically enriched peaks, thus
enabling the assignment of the absolute configurations, as shown
in Figure 1. The oxymethine stereochemistry of substrate olefins
6b and 6c did not affect the enantioselectivity of asymmetric
epoxidations, with the exception of Z-olefin 6d.
conversion
/%
(epoxide)
8a
8b
8c
8d
40 (15 min)
71 (2 h)
8a-1 (cis)
49
51
38
62
33
67
61
39
99 (4R,5R)
87 (4R,5R)
14 (4S)
8a-2 (trans)
8b-1 (cis)a
8b-2 (trans)a
8c-1 (cis)
8c-2 (trans)
8d-1 (cis)a
8d-2 (trans)a
27 (4R)
22 (2 h)
4 (4R,5R)
94 (4R,5R)
76 (4R,5S)
7 (4S,5R)
32 (2 h)
aExchangeable.
buffer.4 GC-MS analysis of the extracts confirmed the production
of 8b-8d and ent-8b-ent-8d from 6b-6d (Figure S2). On the other
hand, no new peak was detected in the reactions with 6e and 6f.
Long incubation times led to decreased conversions, possibly due
to the degradation of the substrates and products by the host strain.
The conversion efficiency decreased in the following order:
trisubstituted 8a {40% (15 min)}, 8b (71%), disubstituted 8d
(32%), and 8c (22%) (Table 1). The enantiofacial selectivity of
nearly all the substrates suggested a re-face attack at the C4
position. In the reactions with very low enantioselectivity (8b-1:
14% ee, 8d-2: 7% ee), reversed enantioselectivity (si-face attack)
Synthetic analogs 6a-6f were separately incubated with
Lsd18-overexpressed Rhodococcus erythropolis in the reaction
1780 | Chem. Lett. 2014, 43, 1779–1781 | doi:10.1246/cl.140721
© 2014 The Chemical Society of Japan

(i)
(ii)
Primary site
Secondary site
Primary site
Secondary site
The epoxidation of 6b, which has the same partial structure as
that of the biosynthetic intermediates of oxasqualenoids, suggested
that Lsd18 epoxidizes trisubstituted olefin that resembles the
isoprene unit. The reaction with (R)-linalool was previously
reported to afford the corresponding epoxides in a nearly 1:1
mixture of the diastereomers.4 The epoxidation of substrate15
mimicking the partial structure of oxasqualenoid afforded a pair
of diastereomers in equal amounts (Figure S3), suggesting that
Lsd18 epoxidizes isoprene-type olefins in a nonenantiofacially
selective manner, unlike polyketide polyenes.
In summary, Lsd18-catalyzed epoxidation was examined
with six differently substituted olefins to evaluate the structural
requirements of Lsd18 substrate recognition. The substitution
pattern on the olefin moiety significantly affected the conversion
and enantiofacial selectivity. Based on the results, a substrate
binding model of epoxidase that explains the enantiofacial
selectivity of various olefins was proposed. The model can be
applied to most enantioselective enzymatic epoxidations of PKS-
derived E-olefins in polyether biosynthesis.
Me
d
d
Me
c
c
b
b
Me
a
180° rotation
of substrate
Me
a
Matched
Mismatched
Figure 2. Substrate binding model of 8a in Lsd18-catalyzed
epoxidation: a: cyclopentyl moiety, b: flexible linker, c: olefin plane,
d: flavin wall.
was observed. These results indicated that Lsd18 binds to the
substrates and exposes the olefin plane to the 4-hydroperoxyflavin
moiety, an oxidation agent in the enzymatic epoxidation.
Substrates harboring C5 methyl groups were converted to the
corresponding epoxides, while no conversion was detected in the
reactions with 6e and 6f, suggesting that substituents in the C5
position are crucial for substrate binding and enantiofacial
selectivity. With regard to the conversion, 5,5-dimethyl-substituted
olefin 6b resulted in a better yield than 5-methyl-substituted olefins
6c and 6d. On the other hand, the observed selectivity in the
reactions with 6c and 6d was better than that with 6b, indicating
that the substitution pattern played a role in the enantiofacial
selectivity. The higher enantiofacial selectivity and conversion of
6a than that of 6c, in addition to the lack of conversion with gem-
disubstituted 6e, indicated that C4-methyl substitution may assist
in fixing the olefin plane, but is not a primary determinant in Lsd18
recognition. Based on these results, a substrate binding model of
Lsd18 was proposed. Lsd18 has primary and secondary recog-
nition sites for C5 and C4 methyl groups, respectively, to fix the
olefin plane of 6a (Figure 2(i)). An internal rotation of the olefin
plane by 180° cannot occur, partly due to the steric hindrance
between the C4 methyl group and Lsd18 (Figure 2(ii)). In the case
of 6b, both orientations can be accepted by Lsd18 with similar
efficiencies because of the lack of the C4 methyl group. Addi-
tionally, this resulted in a good conversion, but low enantiose-
lectivity. The difference in enantioselectivity in the reactions with
6c and 6d was probably due to the C1 hydroxy group. In order to
predict the enantioselectivity of the substrates, the binding model
should be fine-tuned by incorporating the crystal structure data.
The proposed binding model can explain the enantiofacial
This work was supported by a MEXT research grant on
innovative area No. 22108002 to H. Oikawa and No. 26750361 to
A. Minami.
Supporting Information is available electronically on J-STAGE.
References
1
D. E. Cane, W. D. Celmer, J. W. Westley, J. Am. Chem. Soc. 1983, 105,
3594.
2
P. F. Leadlay, J. Staunton, M. Oliynyk, C. Bisang, J. Cortés, E. Frost,
Z. A. Hughes-Thomas, M. A. Jones, S. G. Kendrew, J. B. Lester, P. F.
Long, H. A. I. McArthur, E. L. McCormick, Z. Oliynyk, C. B. W. Stark,
C. J. Wilkinson, J. Ind. Microbiol. Biotechnol. 2001, 27, 360.
Y. Shichijo, A. Migita, H. Oguri, M. Watanabe, T. Tokiwano, K.
Watanabe, H. Oikawa, J. Am. Chem. Soc. 2008, 130, 12230.
A. Minami, M. Shimaya, G. Suzuki, A. Migita, S. S. Shinde, K. Sato, K.
Watanabe, T. Tamura, H. Oguri, H. Oikawa, J. Am. Chem. Soc. 2012,
134, 7246.
3
4
5
6
7
Y. Matsuura, Y. Shichijo, A. Minami, A. Migita, H. Oguri, M. Watanabe,
T. Tokiwano, K. Watanabe, H. Oikawa, Org. Lett. 2010, 12, 2226.
A. Minami, A. Migita, D. Inada, K. Hotta, K. Watanabe, H. Oguri, H.
Oikawa, Org. Lett. 2011, 13, 1638.
K. Hotta, X. Chen, R. S. Paton, A. Minami, H. Li, K. Swaminathan, I. I.
Mathews, K. Watanabe, H. Oikawa, K. N. Houk, C.-Y. Kim, Nature
2012, 483, 355.
8
9
K. Sato, A. Minami, T. Ose, H. Oguri, H. Oikawa, Tetrahedron Lett.
2011, 52, 5277.
A. Minami, T. Ose, K. Sato, A. Oikawa, K. Kuroki, K. Maenaka, H.
Oguri, H. Oikawa, ACS Chem. Biol. 2014, 9, 562.
selectivity of diene precursor
2 in lasalocid biosynthesis
(Figure 1). The trisubstituted E-olefin moiety is a common
structural motif found in most biosynthetic polyene precursors of
ionophore polyethers. Intriguingly, in most cases, the epoxidation
occurs via the re-face to afford (R,R)-epoxides. Thus, the proposed
model can be applied to the epoxidation of trisubstituted E-olefins.
In some cases, the installation of (S,S)-epoxides on disubstituted
E-olefins also occurs in the biosynthesis of several ionophore
polyethers, including monensin (Scheme 1B). In the original
prediction, the substitution pattern (tri- vs. disubstitution) was
deemed to be important in defining the enantiofacial selectivity.
However, changing the substitution pattern from trisubstituted 6a
to disubstituted 6c did not result in the switched enantiofacial
selectivity. One of the possible reasons behind this observation
was that the simplified monoolefin analogs did not have
sufficiently large substituents in the C5 position to mimic the
terminus of triene precursor 5.
10 a) A. Minami, H. Oguri, K. Watanabe, H. Oikawa, Curr. Opin. Chem.
Biol. 2013, 17, 555. b) A. R. Gallimore, Nat. Prod. Rep. 2009, 26, 266.
11 a) A. Migita, M. Watanabe, Y. Hirose, K. Watanabe, T. Tokiwano, H.
Kinashi, H. Oikawa, Biosci., Biotechnol., Biochem. 2009, 73, 169. b) Y.
Sun, X. Zhou, H. Dong, G. Tu, M. Wang, B. Wang, Z. Deng, Chem. Biol.
2003, 10, 431. c) B. M. Harvey, T. Mironenko, Y. Sun, H. Hong, Z. Deng,
P. F. Leadlay, K. J. Weissman, S. F. Haydock, Chem. Biol. 2007, 14, 703.
d) C. Jiang, H. Wang, Q. Kang, J. Liu, L. Bai, Appl. Environ. Microbiol.
2012, 78, 994. e) M. E. Yurkovich, P. A. Tyrakis, H. Hong, Y. Sun, M.
Samborskyy, K. Kamiya, P. F. Leadlay, ChemBioChem 2012, 13, 66.
12 M. Tosin, L. Smith, P. F. Leadlay, Angew. Chem., Int. Ed. 2011, 50,
11930.
13 E. Negishi, S.-Y. Liou, C. Xu, S. Huo, Org. Lett. 2002, 4, 261.
14 O. A. Wong, Y. Shi, Chem. Rev. 2008, 108, 3958.
15 Y. Morimoto, T. Iwai, T. Yoshimura, T. Kinoshita, Bioorg. Med. Chem.
Lett. 1998, 8, 2005.
Chem. Lett. 2014, 43, 1779–1781 | doi:10.1246/cl.140721
© 2014 The Chemical Society of Japan | 1781