Identification of a nonaketide product for the iterative polyketide synthase in biosynthesis of the nine-membered enediyne C-1027

Article abstract of DOI:10.1002/anie.201003369

Not octa- but nonaketide: The nine-membered enediyne core polyketide synthase SgcE efficiently synthesizes a nonaketide in the absence of any assisting proteins (see scheme), contrary to the suggestion that an octaketide is the product of the synthase under assistance from a thioesterase. This finding redefines the catalytic functions of the polyketide synthase. CoA=coenzyme A.

Full text of DOI:10.1002/anie.201003369

Communications
DOI: 10.1002/anie.201003369
Enediyne Biosynthesis
Identification of a Nonaketide Product for the Iterative Polyketide
Synthase in Biosynthesis of the Nine-Membered Enediyne C-1027**
Xiaolei Chen, Zu-Feng Guo, Pok Man Lai, Kong Hung Sze,* and Zhihong Guo*
Enediyne natural products are potent antitumor antibiotics
characterized by an enediyne core containing two acetylenic
groups conjugated to a double bond in a nine- or ten-
membered carbocycle.[1] Their structural complexity poses a
formidable hurdle to the elucidation of their biosynthetic
origin. In recent years, studies from the Shen and Thorson
research groups have revealed a unified paradigm for
enediyne biosynthesis,[2–6] which involves a conserved iter-
ative type I polyketide synthase (PKS) responsible for syn-
thesis of the carbon skeleton of the enediyne cores.[7–9]
However, most enzymatic steps involved in assembly of the
enediyne cores remain unknown.
Characterization of the conserved iterative PKS enzymes
has led to different perspectives on enediyne core biosynthe-
sis. Shen et al. isolated an all-trans heptaene 1 (Figure 1A)
from heterologous cells expressing the PKS SgcE from the
gene cluster of C-1027 and proposed it as a potential
precursor for the nine-membered enediyne core.[10] Subse-
quently, Liang et al. identified a 1-resembling hexaenone as
the product of the PKS CalE8 from the gene cluster of the
ten-membered calicheamicin,[11] but later suggested heptaene
1 as the common precursor of both nine- and ten-membered
enediyne cores in a separate study.[12] However, recharacte-
rization of CalE8 by Townsend et al. suggests that neither
heptaene 1 nor the hexaenone is a true precursor of
calicheamicin and that a trans-acting enzyme may be
needed for proper functioning of the PKS.[13]
Similarly, a more recent in vivo functional investigation
suggests a universal PKS-bound intermediate for modifica-
tion by other enzymes en route to the enediyne cores.[14] In
these studies, the polyene-forming activity of the PKSs was
found to be strictly dependent on a conserved hot-dog
Figure 1. A) Domain organization of the enediyne PKS SgcE and its
previously proposed involvement in the biosynthesis of C-1027. ACP,
acyl carrier protein; PPTase, phosphopantetheinyl transferase; see text
[*] X. Chen, Dr. Z.-F. Guo, Prof. Z. Guo
for definitions of KS, AT, KR, and DH. B) HPLC and UV/Vis analyses of
Department of Chemistry, State Key Laboratory of Molecular
the ethyl acetate extracts from enzymic reactions containing 2.5 mm
Neuroscience
SgcE at pH 7.0. The HPLC chromatograph and UV/Vis spectrum of the
The Hong Kong University of Science and Technology
same extract are given in the same color. The UV/Vis detector was set
Clear Water Bay, Kowloon, Hong Kong SAR (China)
at 355 nm in the HPLC analyses. C) UV/Vis spectra of the products 1–
Fax: (+852)2358-1594
3.
E-mail: chguo@ust.hk
Homepage: http://chem.ust.hk/Faculty%20staff/Guo/content.htm
P. M. Lai, Prof. K. H. Sze
Department of Chemistry Open Laboratory of Chemical Biology of
the Institute of Molecular Technology for Drug Discovery and
without reliance on the thioesterase, which suggests a possible
Synthesis, The University of Hong Kong
Pokfulam Road, Hong Kong SAR (China)
thioesterase. Herein, we report that SgcE synthesizes a
nonaketide product distinct from all the reported polyenes
new biosynthetic route to the enediyne cores.
SgcE was both expressed alone and co-expressed with the
hot-dog-fold thioesterase SgcE10 in Escherichia coli. Hep-
taene 1 was readily isolated from cells co-expressing SgcE and
SgcE10 (see the Supporting Information), but was not
detected in cells expressing SgcE only. Moreover, the isolated
E-mail: khsze@hku.hk
[**] We thank the Research Grants Council of the HKSAR government
for financial support (RPC10SC07).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201003369.
7926
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7926 –7928

Angewandte
Chemie
SgcE was yellowish with a broad absorption from 350 to
500 nm when it was expressed alone, but was essentially
colorless when it was co-expressed with the thioesterase.
These observations are consistent with the results in the
previous in vivo functional characterization of SgcE.[10]
Both the yellowish and colorless forms of SgcE are active
in assays at pH 7.0 using acetyl-coenzyme A (CoA), malonyl-
CoA, and reduced nicotinamide adenine dinucleotide phos-
phate (NADPH) as substrates. Extraction of the reaction
mixture by ethyl acetate affords a yellowish solution exhibit-
ing multipeak UV/Vis absorptions, which overlap well with
most parts of the absorption spectrum of heptaene 1 except
for the most intense peak at 394 nm (Figure 1B). Only one
product (2) with essentially the same optical absorptions is
detected in the extract by HPLC methods (Figure 1C), thus
indicating that it is the only extractable, chromogenic product
of the enzymatic reaction.
This outcome is not changed by inclusion of the thioester-
ase SgcE10 in the enzymatic reaction at a concentration up to
2 mm. However, further increase of the SgcE10 concentration
in the enzymatic reaction steadily decreases the yield of
product 2 and leads to the production with increasing yields of
two additional products, 3 and a compound with a retention
time of 27 minutes (Figure 1B); the latter was confirmed to
be identical to heptaene 1 by HPLC co-injection and their
identical UV/Vis spectra. Similar results were obtained for
reactions at pH 8.0, in which the products 3 and 4 were
formed with higher yields at lower SgcE10 concentrations
(see the Supporting Information, Figure S3). Products 1 and 3
cannot be formed from product 2 in the presence of SgcE10.
Product 2 was isotope-labeled using 13C-malonic acid and
determined to be a nonaketide, 5,6-dihydro-4-hydroxy-6-
(1E,3E,5E,7E,9E,11E-tridecahexaen-1-yl)-2H-pyran-2-one,
by various NMR techniques in combination with mass
spectrometric analysis of the turnover product before and
after palladium-catalyzed hydrogenation. Product 2 exists in
equilibrium between a major enol tautomer (Scheme 1) and a
minor keto tautomer. Meanwhile, product 3 was determined
to be 2-oxo-5E,7E,9E,11E,13E,15E-heptadecahexaen-4-ol by
mass spectrometry and NMR techniques. It contains the same
p-conjugation system as product 2, consistent with the
observation of identical UV/Vis spectra for these two com-
pounds (Figure 1C). Notably, the six conjugated double
bonds in both products are all in the trans configuration,
since J-resolved 2D NMR experiments show that the splitting
of the olefinic proton signals is closely similar to that of
heptaene 1, the hexaenone congener of which has been
confirmed to contain only trans-configured double bonds.[12]
The structure of product 2 allows us to deduce the
biosynthetic logic of the iterative type I PKS SgcE. After
being primed with an acetyl group as the starter unit, the PKS
catalyzes eight consecutive decarboxylative Claisen conden-
sations of malonyl CoA thioester. The b-ketoacyl synthase
(KS), acyl transferase (AT), ketoreductase (KR), and dehy-
dratase (DH) domains of the synthase function sequentially
in each of the first six repetitive polyketide chain elongation
cycles to form the six conjugated double bonds. Subsequently,
the DH domain loses its catalytic activity in the last two
elongation cycles and the KR domain is inactive in the last
Scheme 1. Proposed catalytic functions for SgcE and SgcE10 in biosyn-
thesis of the enediyne antibiotic C-1027.
cycle, probably because of the unique chain-length control
mechanism of the PKS, thus leading to the formation of 3-
oxo-5-hydroxy-6E,8E,10E,12E,14E,16E-octadecahexaenoyl-
S-SgcE. Finally, a thermodynamically favorable d lactoniza-
tion affords the nonaketide product 2 (Scheme 1). This SgcE
catalytic process does not require the assistance of a second
enzyme and is not affected by SgcE10 at a low to moderate
concentration, contrary to the suggestion that SgcE10 is
indispensable to SgcE activities.[10,12]
The origins of products 1 and 3 are also implicated in their
structures. These products are formed from hydrolysis of the
SgcE-bound intermediates by SgcE10 and subsequent decar-
boxylation of the b-oxo-carboxylates (Scheme 1). As such,
both compounds are aberrant products formed only in the
presence of the thioesterase at high concentrations. It is not
clear why the intermediates giving rise to 1 and 3 are more
susceptible than others to hydrolysis by SgcE10 at pH 7.0.
However, other intermediates in the polyketide synthesis
appear to be also accessible by SgcE10 to form more side
products at pH 8.0 (see the Supporting Information, Fig-
ure S3).
The behavior of SgcE10 in the synthesis of nonaketide 2 is
typical of a type II thioesterase commonly found in poly-
ketide and nonribosomal peptide syntheses.[15,16] The thioes-
terase is thus likely to be a proofreading enzyme to regenerate
the mis-acylated PKS under physiological conditions. Its
aberrant activity to hydrolyze the normal biosynthetic
intermediates at high concentrations is consistent with the
finding of heptaene 1 in cells co-expressing both SgcE and
SgcE10 but not in cells expressing SgcE only.[10] The absence
of products 2 and 3 in SgcE-expressing cells may be because
of their degradation by host enzymes. Moreover, the failure to
detect products 2 and 3 in a previous HPLC analysis of the
Angew. Chem. Int. Ed. 2010, 49, 7926 –7928
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7927

Communications
SgcE products may result from the setting of the UV/Vis
Keywords: biosynthesis · enediynes · enzyme catalysis ·
natural products · polyketides
.
detector at 400 nm,[12] at which neither of these products
shows absorption.
Product 2 is structurally distinct from the minor pyrone
side products formed from the short-chain biosynthetic
intermediates en route to the mature polyketide product.[13,17]
As the only molecule synthesized efficiently at a physiological
pH value, it is a potential precursor in the biosynthesis of the
C-1027 enediyne core. This finding suggests a probable new
biosynthetic route to the nine-membered enediyne core of C-
1027, although it is not immediately clear how the four extra
carbon atoms in the nonaketide are shed in the process. Since
the enediyne PKSs are interchangeable within the nine-
membered family,[10] this newly found nonaketide is also a
potential precursor for other nine-membered enediyne cores.
This likely role of the nonaketide does not contradict the
isotope-labeling pattern of neocarzinostatin.[18] To provide
direct evidence for this potential role, product 2 was isotope-
labeled for incorporation into the final enediyne natural
product. However, this attempt failed probably because of
instability of the nonaketide under cell-culturing conditions,
consistent with the inability to find this compound in cells
expressing the PKS.
In summary, we have redefined SgcE as a specific
producer of nonaketide 2, thus demonstrating that it is an
iterative PKS with a distinct chain-termination mechanism in
which its functional domains are sequentially disengaged
from the polyketide synthesis. In the meantime, we have also
shown that the hot-dog protein SgcE10 is most likely a type II
thioesterase in the polyketide synthesis. These findings may
facilitate elucidation of the intriguing biosynthetic steps
leading to the nine-membered enediyne cores. In addition,
the clear distinction between this nonaketide and the
octaketide polyene products reported for PKSs from the
ten-membered enediyne family raises the question again of
whether biosynthesis of nine- and ten-membered enediynes
diverges at the PKS stage.
[1] A. L. Smith, K. C. Nicolaou, J. Med. Chem. 1996, 39, 2103 – 2117.
[2] W. Liu, S. D. Christenson, S. Standage, B. Shen, Science 2002,
297, 1170 – 1173.
[3] J. Ahlert, E. Shepard, N. Lomovskaya, E. Zazopoulos, A. Staffa,
B. O. Bachmann, K. Huang, L. Fonstein, A. Czisny, R. E.
Whitwam, C. M. Farnet, J. S. Thorson, Science 2002, 297, 1173 –
1176.
[4] W. Liu, K. Nonaka, L. Nie, J. Zhang, S. D. Christenson, J. Bae,
S. G. Van Lanen, E. Zazopoulos, C. M. Farnet, C. F. Yang, B.
Shen, Chem. Biol. 2005, 12, 293 – 302.
[5] S. G. Van Lanen, T. Oh, W. Liu, E. Wendt-Pienkowski, B. Shen,
J. Am. Chem. Soc. 2007, 129, 13082 – 13094.
[6] Q. Gao, J. S. Thorson, FEMS Microbiol. Lett. 2008, 282, 105 –
114.
[7] W. Liu, J. Ahlert, Q. Gao, E. Wendt-Pienkowski, B. Shen, J. S.
Thorson, Proc. Natl. Acad. Sci. USA 2003, 100, 11959 – 11963.
[8] E. Zazopoulos, K. Huang, A. Staffa, W. Liu, B. O. Bachmann, K.
Nonaka, J. Ahlert, J. S. Thorson, B. Shen, C. M. Farnet, Nat.
Biotechnol. 2003, 21, 187 – 190.
[9] G. P. Horsman, S. G. Van Lanen, B. Shen, Methods Enzymol.
2009, 459, 97 – 112.
[10] J. Zhang, S. G. Van Lanen, J. Ju, W. Liu, P. C. Dorrestein, W. Li,
N. L. Kelleher, B. Shen, Proc. Natl. Acad. Sci. USA 2008, 105,
1460 – 1465.
[11] R. Kong, L. P. Goh, C. W. Liew, Q. S. Ho, E. Murugan, B. Li, K.
Tang, Z.-X. Liang, J. Am. Chem. Soc. 2008, 130, 8142 – 8143.
[12] H. Sun, R. Kong, D. Zhu, M. Lu, Q. Ji, C. W. Liew, J. Lescar, G.
Zhong, Z.-X. Liang, Chem. Commun. 2009, 7399 – 7401.
[13] K. Belecki, J. M. Crawford, C. A. Townsend, J. Am. Chem. Soc.
2009, 131, 12564 – 12566.
[14] G. P. Horsman, Y. Chen, J. S. Thorson, B. Shen, Proc. Natl. Acad.
Sci. USA 2010, 107, 11331 – 11335.
[15] B. S. Kim, T. A. Cropp, B. J. Beck, D. H. Sherman, K. A.
Reynolds, J. Biol. Chem. 2002, 277, 48028 – 48034.
[16] Z.-F. Guo, Y. Sun, S. Zheng, Z. Guo, Biochemistry 2009, 48,
1712 – 1722.
[17] Z.-X. Liang, Nat. Prod. Rep. 2010, 27, 499 – 528.
[18] O. D. Hensens, J. L. Giner, I. H. Goldberg, J. Am. Chem. Soc.
1989, 111, 3295 – 3299.
Received: June 3, 2010
Published online: September 16, 2010
7928 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7926 –7928