A R T I C L E S
Lin et al.
1-phenyl-1,2-ethanediol was produced with 20% ee (Panel
VI in Figure 6). This labeling pattern is consistent with a
lack of regiospecificity of the enzyme in attacking both C-1
and C-2 of (R)-styrene oxide (Figure 7B). These data establish
that the enantiomerically enriched (R)-vicinal diol product
from SgcF-catalyzed hydrolysis of racemic styrene oxide
originates from complementary enantio- and regioselectivities
of SgcF.
vivo data showed that no C-1027 production was observed
in the ∆sgcF mutant strain SB1010, unambiguously estab-
lishing that sgcF is indispensable for C-1027 biosynthesis.
In vitro characterization of SgcF unveiled a canonical EH
activity as demonstrated by the efficient hydrolysis of styrene
oxide, a substrate mimic of the putative enediyne core
epoxide (Figure 3). Specifically, SgcF efficiently hydrolyzes
(S)-styrene oxide, displaying an apparent Km of 0.6 ( 0.1
mM and kcat of 48 ( 1 min-1, via attack at the R-position to
exclusively generate the (R)-phenyl vicinal diol, consistent
with the stereochemistry of the C-1027 chromophore. These
findings support the role of SgcF in the proposed convergent
pathway for C-1027 biosynthesis, unveiling an (R)-vicinal
diol as a key intermediate (Figures 1 and 3). Interestingly,
SgcF can also hydrolyze (R)-styrene oxide to afford prefer-
entially the (R)-phenyl vicinal diol via attack at the ꢀ-position,
albeit with significantly reduced efficiency (apparent Km of
2.0 ( 0.4 mM and kcat ) 4.3 ( 0.3 min-1). However, the
latter activity unlikely contributes to C-1027 biosynthesis in
vivo since the oxygenase that catalyzes the formation of the
enediyne core epoxide intermediate most probably would
yield a single enediyne core epoxide, that is, the (S)-
enantiomer, as a substrate for SgcF.
EHs are believed to be involved in the biosynthesis of a
variety of natural products, including polyether antibiotics like
nanchangmycin60 and monensin,61 plant cutin,29 and insect
pheromones.62 The large diversity of putative EHs identified
from genome sequences49 implies that comparative studies may
potentially reveal a great deal about the determinants of regio-
and enantiospecificity of these enzymes. Indeed, comparative
analyses of enediyne biosynthetic gene clusters33-35 has pre-
dicted two EHs involved in the production of NCS - NcsF1
(53% identity to SgcF) and NcsF2 (62% identity to SgcF).34 In
a biosynthetic strategy that parallels SgcF, one would predict
NcsF1 or NcsF2 to catalyze the hydrolysis of an (R)-enediyne
core epoxide intermediate to afford an (S)-vicinal diol product
for NCS biosynthesis (Figure 1). Comparative studies of SgcF
and NcsF1 or F2 therefore would present a great opportunity
to investigate how EHs evolve and acquire their exquisite regio-
and enantioselectivity.
Discussion
C-1027, a chromoprotein antitumor antibiotic produced by
Steptomyces globisporus, is composed of an apoprotein (CagA)
and the C-1027 chromophore. The C-1027 chromophore consists
of four distinct moieties - an enediyne core, a deoxy aminosugar,
a benzoxazolinate, and an (S)-3-chloro-5-hydroxy-ꢀ-tyrosine
moiety (Figure 1).50-52 The 9-membered enediyne core readily
undergoes Bergman cycloaromatization to generate a highly
reactive diradical intermediate that abstracts hydrogen atoms
from DNA, leading to both double-stranded breaks (DSBs) and
interstrand cross-links (ICLs) that ultimately result in cell
death.53-55 Although the enediyne core directly confers cyto-
toxicity, the three peripheral moieties appended to the core are
required for activity and stability. For instance, the benzoxazo-
linate moiety aids in binding to CagA that protects and carries
the enediyne chromophore,56,57 while the (S)-3-chloro-5-hy-
droxy-ꢀ-tyrosine also contributes critical binding interactions
with CagA58,59 and modulates the reactivity of the enediyne
core via π-π interactions.59 A comparison of the biosynthetic
gene clusters of C-1027,33 NCS34 and maduropeptin35 suggested
that these two moieties might be appended to a (R)-vicinal diol-
containing C-1027 enediyne core, which in turn may originate
from a (S)-enediyne core epoxide intermediate via EH-catalyzed
ring-opening (Figure 3).
To test this hypothesis, we first carried out extensive
bioinformatics analysis of the 56 genes within C-1027
biosynthetic gene cluster, and identified two genes, sgcF and
sgcI, encoding R/ꢀ- hydrolases.33 While bioinformatics
analysis alone fell short of assigning SgcI as an EH, SgcF
possesses such EH characteristics as the nucleophilic elbow
(Sm-X-Nu-X-Sm-Sm), the G-X-Sm-X-S/T motif, the H-G-
X-P oxyanion hole, and the Asp-His-Asp catalytic triad, and
hence was annotated as an EH. Interestingly, SgcF contains
only one of the two Tyr residues usually conserved in the
lid domain (i.e., Y304) that serve to activate the oxirane ring
to nucleophilic attack, and the other is changed to Trp (i.e.,
W236) (Figure S2, Supporting Information).10,15-17,20-22,26 In
Most known EHs catalyze the hydrolysis of each enantiomer
of a racemic mixture at a different rate to yield enantiomerically
enriched vicinal diol products. However, this method of
separation, or kinetic resolution, is limited by a maximum
possible yield of 50%. In contrast, SgcF can not only catalyze
the hydrolysis of (S)-styrene oxide to afford (R)-1-phenyl-1,
2-ethanediol exclusively but also hydrolyze (R)-styrene oxide
to yield predominately (R)-1-phenyl-1, 2-ethanediol with 20%
ee. Taken together, SgcF can generate (R)-1-phenyl-1, 2-ethanediol
from racemic styrene epoxide with a minimal enantiomeric
excess of 60% ee. SgcF therefore represents an excellent starting
point for protein engineering efforts directed toward developing
a biocatalyst for preparing enantiomerically pure diols from
racemic epoxides.
(50) Hu, J.; Xue, Y. C.; Xie, M. Y.; Zhang, R.; Otani, T.; Minami, Y.;
Yamada, Y.; Marunaka, T. J. Antibiot. 1988, 41, 1575–1579.
(51) Otani, T.; Minami, Y.; Marunaka, T.; Zhang, R.; Xie, M. Y. J. Antibiot.
1988, 41, 1580–1585.
(52) Otani, T.; Yasuhara, T.; Minami, Y.; Shimazu, T.; Zhang, R.; Xie,
M. Y. Agric. Biol. Chem. 1991, 55, 407–417.
(53) Kennedy, D. R.; Gawron, L. S.; Ju, J. H.; Liu, W.; Shen, B.; Beerman,
T. A. Cancer Res. 2007, 67, 773–781.
(54) Kennedy, D. R.; Ju, J.; Shen, B.; Beerman, T. A. Proc. Natl. Acad.
Sci. U.S.A. 2007, 104, 17632–17637.
In conclusion, the in vitro characterization of SgcF as an EH
significantly advances our understanding of C-1027 biosynthesis,
(55) Xu, Y. J.; Zhen, Y. S.; Goldberg, I. H. Biochemistry 1994, 33, 5947–
5954.
(56) Matsumoto, T.; Okuno, Y.; Sugiura, Y. Biochem. Biophys. Res.
Commun. 1993, 195, 659–666.
(60) Sun, Y.; Zhou, X.; Dong, H.; Tu, G.; Wang, M.; Wang, B.; ZX, D.
Chem. Biol. 2003, 10, 431–441.
(57) Okuno, Y.; Otsuka, M.; Sugiura, Y. J. Med. Chem. 1994, 37, 2266–
2273.
(61) Gallimore, A.; Stark, C.; Bhatt, A.; Harvey, B.; Demydchuk, Y.;
Bolanos-Garcia, V.; Fowler, D.; Staunton, J.; Leadlay, P.; Spencer, J.
Chem. Biol. 2006, 13, 453–460.
(58) Okuno, Y.; Iwashita, T.; Sugiura, Y. J. Am. Chem. Soc. 2000, 122,
6848–6854.
(59) Tanaka, T.; Fukuda-Ishisaka, S.; Hirama, M.; Otani, T. J. Mol. Biol.
2001, 309, 267–283.
(62) Abdel-Latief, M.; Garbe, L. A.; Koch, M.; Ruther, J. Proc. Natl. Acad.
Sci. U.S.A. 2008, 105, 8914–8919.
9
16416 J. AM. CHEM. SOC. VOL. 131, NO. 45, 2009