.
Angewandte
Communications
DOI: 10.1002/anie.201305569
Combinatorial Biosynthesis
Multiplexing of Combinatorial Chemistry in Antimycin Biosynthesis:
Expansion of Molecular Diversity and Utility**
Yan Yan, Jing Chen, Lihan Zhang, Qingfei Zheng, Ying Han, Hua Zhang, Daozhong Zhang,
Takayoshi Awakawa, Ikuro Abe, and Wen Liu*
The control of biological phenomena by natural products,
which have co-evolved with their macromolecular receptors,
results in their critical role in medicinal chemistry and
chemical biology.[1] The generation of libraries of natural-
product-like compounds is a major area of current interest in
diversity-oriented synthesis, however, synthetic accessibility
and efficiency of these compounds may face challenges
associated with their structural complexity.[1] Natural prod-
ucts that include polyketides, nonribosomal peptides, and
their hybrids often share a similar biosynthetic logic,[2]
featuring the preparation of building blocks, skeleton assem-
bly catalyzed by polyketide synthase (PKS) or/and non-
ribosomal peptide synthetase (NRPS), and post-tailoring for
molecule maturation. The formation of antimycins (ANTs),
which display a remarkable variety of biological activities,[3] is
not an exception (Figure 1).
unit, provided by a crotonyl-CoA reductase/carboxylase
(CCR)-like protein AntE,[6] acylation at C8 is the result of
post-NRPS/PKS modification of the dilactone core, mediated
by a promiscuous acyltransferase AntB. The ANT NRPS/PKS
system was also found to tolerate the change of the starter
unit and to initiate assembly of a dilactone conjugated with
a modified 3-formamidosalicyclic acid (FSA) moiety.[6b–d]
Interesting questions thus arise regarding 1) how promiscuous
the entire ANT biosynthetic machinery is and 2) whether
multiple alterations can be achieved to create unnatural
ANTs. By addressing these queries in the ANT-producing
strain Streptomyces sp. NRRL 2288, we herein report a multi-
plex combinatorial biosynthesis approach, to expand the
molecular diversity and utility upon applying an idea of
combinatorial chemistry to different biosynthetic stages of
ANTs.
ANTs have a nine-membered polyketide–peptide hybrid
dilactone proven to be a privileged scaffold capable of
binding multiple protein targets.[3] We and others have
uncovered a uniform paradigm for their biosynthesis,[4]
showing that a hybrid NRPS/PKS system programs the
formation of the dilactone core in a linear way (Figure 1).
The 44 natural ANTs differ in their alkylation at C7 and
acylation at C8,[5] however, the effects of these functionalities
on biological activities and associated modes of action remain
unclear. Very recently, Zhang et al.[4d,e] established the
biochemical basis for these two modifications (Figure 1):
while alkylation at C7 occurs through PKS-catalyzed elonga-
tion that utilizes variable alkylmalonyl units as the extender
We first deleted antB, aiming at the construction of an
engineered biosynthetic apparatus for diversity-oriented
production of dilactone scaffolds in vivo. The resulting
mutant strain, AL2110, failed to produce mature ANTs but
accumulated a series of C8-deacylated ANTs that vary in the
alkylation at C7, including DA-1, DA-2, DA-3, DA-4, and
DA-5 (Figures 2 and S2). Three carboxylates, chloropenta-
noate (1), cyclohexanepropanoate (2), and 10-undecynoate
(3), were then fed to AL2110, to examine whether naturally
unavailable units can be incorporated into ANTs to increase
the diversity at C7. These acids are known to be activated in
cells by endogenous acyl-CoA ligase(s), giving chloropenta-
noyl-CoA (4), cyclohexanepropanoyl-CoA (5), and 10-unde-
cynoyl-CoA (6), respectively, which go further through
a partial b-oxidation pathway to generate a,b-unsaturated
products, chloropentenoyl-CoA (7), cyclohexanepropenoyl-
CoA (8), and (2E)-2-undecen-10-ynoyl-CoA (9), respectively
(Figure 3a). As a result, all feedings produced new C8-
deacylated ANTs, DA-6–DA-9, each of which became the
major product in the chemical profile of AL2110 (Figure S2).
The variation at C7 is well in line with the carboxylate
precursors, as 3-chloropropyl of DA-6 from 1, cyclohexyl-
methyl of DA-7 from 2, and both 4-pentynyl of DA-8 and 6-
heptynyl of DA-9 from 3 (Figure 2). Synthesis of 4-pentynyl
in DA-8 or 6-heptynyl in DA-9 indicates that the CoA
derivative 9 (eleven carbon atoms) was subjected to one or
two complete round(s) of b-oxidation to produce the short-
[*] Y. Yan, Dr. J. Chen, Q. Zheng, Y. Han, H. Zhang, D. Zhang,
Prof. Dr. W. Liu
State Key Laboratory of Bioorganic and Natural Products Chemistry
Shanghai Institute of Organic Chemistry (SIOC)
Chinese Academy of Sciences (CAS)
345 Lingling Road, Shanghai 200032 (China)
E-mail: wliu@mail.sioc.ac.cn
Dr. J. Chen, L. Zhang, Prof. Dr. T. Awakawa, Prof. Dr. I. Abe
Graduate School of Pharmaceutical Sciences, University of Tokyo
7-3-1 Hongo, Bunkyo-ku (Japan)
[**] We thank Yu Cai at SIOC, CAS, and Prof. Junying Yuan at Harward
Medical School for providing HeLa cells and assistance in cell-
labeling experiments. This work was supported in part by grants
from the NNSF (91213303), STCSM (13XD1404500), and “973
program” (2010CB833200 and 2012CB721100) of China (for W.L.),
and by Grants-in-Aid for Scientific Research from the Ministry of
Education, Culture, Sports, Science, and Technology of Japan (for
I.A.).
À
ened derivatives (2E)-2-nonen-8-ynoyl CoA (14, 2C) or
À
(2E)-2-hepten-6-ynoyl CoA (15, 4C; Figure 3a). For poly-
ketide extension, in situ reductive carboxylation at C2 of 7, 8,
14, and 15 requires the activity of the CCR-like protein AntE
to generate the corresponding extender units 10, 11, 12, and
13, respectively, consistent with its surprising promiscuity
Supporting information for this article is available on the WWW
12308
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 12308 –12312