A new mechanism for benzopyrone formation in aromatic polyketide biosynthesis

Article abstract of DOI:10.1021/ja0736919

Aromatic polyketides are all biosynthesized from highly reactive poly-β-ketone intermediates. Keys to introducing the vast chemical diversity seen in these natural products are the enzymatic and non-enzymatic tailoring chemistries that occur after biosynt

Full text of DOI:10.1021/ja0736919

Published on Web 07/11/2007
A New Mechanism for Benzopyrone Formation in Aromatic Polyketide
Biosynthesis
Wenjun Zhang,† Burkhardt I. Wilke,‡ Jixun Zhan,† Kenji Watanabe,§ Christopher N. Boddy,*,‡ and
Yi Tang*,†
Department of Chemical and Biomolecular Engineering, UniVersity of California at Los Angeles, Los Angeles,
California 90095, Department of Chemistry, Syracuse UniVersity, Syracuse, New York 13244, and Department of
Pharmaceutical Sciences, School of Pharmacy, UniVersity of Southern California, Los Angeles, California 90089
Received May 22, 2007; E-mail: cnboddy@syr.edu; yitang@ucla.edu
Aromatic polyketides, an important class of pharmaceutical
agents, are all biosynthesized from highly reactive poly-â-keto acid
intermediates. Keys to introducing the vast chemical diversity seen
in these natural products are the enzymatic and non-enzymatic
tailoring chemistries that occur after biosynthesis of the poly-â-
keto backbone. In this work, we expand the scope of non-enzyme-
catalyzed modifications and show that primary amides can act in
vivo as electrophiles, facilitating the formation of benzopyrones.
We demonstrate this mechanism can be rationally introduced into
an engineered biosynthetic pathway to produce new compounds.
This mechanism is of particular note since it demonstrates the use
of a “protecting group” in polyketide biosynthesis.
There are two known fates for a primary amide functional group
during aromatic polyketide biosynthesis. The amide can remain
unreacted through the biosynthetic pathway, as seen with tetracy-
cline.1 Alternatively, primary amides can be transformed to pyri-
dones, as is seen in the natural products lysolipin,2 xantholipin,3
and fredericamycin A.4 Our previous work with the oxytetracycline
minimal polyketide synthase (PKS) has shown that pyridone
formation is non-enzymatic.5,6 Expression of the minimal PKS in
conjunction with amidotransferase OxyD in Streptomyces coelicolor
CH999 generates the decaketide backbone 2 (Scheme 1). In the
absence of tailoring enzymes, the novel isoquinoline compound 6
(WJ85) was produced.6 Addition of the C-9 specific ketoreductase
OxyJ led to production of 10 (WJ35).5 The alkaloid-like benzopy-
ridone structures observed in 6 and 10 are derived from a
spontaneous nucleophilic attack of the amide group on a proximal
backbone carbonyl, showing the importance of non-enzyme-
catalyzed reactivity of the amide group in aromatic polyketide
tailoring.
The unique tailoring chemistry observed from the poly-â-keto
amide intermediate 2 prompted us to explore the biosynthesis of
other nitrogen-containing polyketides through coexpression of
additional tailoring enzymes. Coexpression of the bifunctional
cyclase/dehydratase OxyK with the minimal PKS, OxyJ, and OxyD
in CH999 afforded a new metabolite 15 (WJ78) in exceptionally
high yield (150 mg/L). Surprisingly, high-resolution mass spec-
trometry indicated a molecular formula of C19H12O7 (m/z )
375.0462 [M + Na]+, ∆ ) 1.9 mmu), lacking the anticipated
nitrogen atom. Extensive one- and two-dimensional NMR charac-
terizations were performed to reveal that 15 is a novel dibenzopy-
rone as shown in Scheme 1. Confirmation of the ester connectivity
in 15 was obtained by comparison of the key phenolic 13C NMR
signal (δC-11 ) 151 ppm) of 15 to synthetic dibenzopyrone and
dibenzopyridone standards (Supporting Information). The structure
of the dibenzopyrone portion of 15 is related to the well-known
fungal mycotoxin alternariol7,8 and graphislactones.9
We hypothesized that the unexpected pyrone formation in 15
occurred via non-enzyme-catalyzed nucleophilic attack of the C-11
phenol in 14 on the amide carbonyl (Scheme 1). Intermediate 14
was formed from 7 via OxyK-catalyzed C-7/C-12 cyclization,
dehydration of the first ring, and spontaneous C-13/C-18 cyclization.
The intramolecular nucleophilic attack of the C-11 phenol on the
amide, forming the six-membered aromatic lactone, is enthalpically
favorable and is likely to proceed under in vivo conditions.
Intermediate 12 is a key branch point in this pathway and can be
processed via two pathways to produce either dibenzopyrone 15
or the benzopyridone analogue. Rapid OxyK-catalyzed dehydration
of the first ring converts the electrophilic C-11 ketone in 12 into a
nucleophilic phenol, enabling dibenzopyrone formation. If the
dehydration activity of OxyK is slowed, spontaneous nucleophilic
attack of the amide on the C-11 keto group of 12 can occur,
generating the corresponding benzopyridone. This alternate route
is analogous to the mechanism for formation of 6 and 10.
The lack of nitrogen in 15 suggests that the amidotransferase
OxyD may not be required for formation of 15. To test this
hypothesis, the minimal PKS, OxyJ, and OxyK were expressed in
S. coelicolor CH999. This strain did not produce any detectable
levels of 15, indicating the essential role of the amidotransferase
in the biosynthesis of 15. The role of the amide group in the
biosynthesis of 15 is akin to a protecting group as used in synthetic
organic chemistry. By masking the terminal carboxylate group of
the poly-â-keto chain as an amide, spontaneous decarboxylation
of C-19 during polyketide processing is prevented. The amide is
then removed once decarboxylation is no longer problematic.
To demonstrate that benzopyrone formation is spontaneous under
physiological conditions, a simplified model of compound 15 was
synthesized (22, Scheme 2). Construction of the biphenyl core
occurred via Suzuki coupling of the commercially available phenyl
boronic acid 16 and the known arylbromide 17.8 Pinnick oxidation
of the aldehyde gave carboxylic acid 19. Attempts to directly
convert the acid into amide 20 via formation of an activated ester
were unsuccessful, leading to exclusive formation of lactone 21.
The unanticipated benzyl deprotection is likely due to activation
of the aryl benzyl ether oxygen followed by nucleophilic cleavage
of the benzyl group.10 Amide 20 was generated through a two-step
process, avoiding activation of the carbonyl. Esterification of 19
followed by treatment with BuLi and anhydrous NH3 gave the key
amide 20. To model the aqueous environment in vivo, the benzyl
group in 20 was removed with Raney-Ni in aqueous conditions
(pH 7.4) to generate the free phenol, which quantitatively displaced
the amide, leading to production of benzopyrone 21. Deprotection
of the methoxy groups with BBr3 afforded the final product 22.
† University of California at Los Angeles.
‡ Syracuse University.
§ University of Southern California.
9
9304
J. AM. CHEM. SOC. 2007, 129, 9304-9305
10.1021/ja0736919 CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
Scheme 1. Biosynthetic Polyketides Derived from Amidated Backbone
Scheme 2. Dibenzopyrone Synthesis via Amide Displacement
widely present in natural products and are typically formed through
esterification of carboxylic acids, intramolecular hydrolysis of
enzyme-attached thioesters, or through recently reported oxidative
rearrangement.13 The observed involvement of the amide moiety
in pyrone formation may be the mechanism by which pyrones are
formed in other aromatic polyketides, including those in the
rubromycin family. Interestingly, OxyD-like amidotransferase ho-
mologues are present in these gene clusters, while no nitrogen atoms
were present in the reported polyketide structures.4
Acknowledgment. This work is supported by a UC CRCC
grant, NSF CBET #0545860, Syracuse University, and Shimadzu
Scientific. We thank Prof. Mike Jung for helpful discussions.
This synthetic sequence confirms that the amide group can act as
an electrophilic center and be attacked by an intramolecular phenol
to generate dibenzopyrone under neutral, low temperature aqueous
conditions, further supporting the proposed biosynthetic scheme
for 15.
Supporting Information Available: Experimental details, NMR
spectroscopy data, and plasmid construction. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) Chopra, I.; Roberts, M. Microbiol. Mol. Biol. ReV. 2001, 65, 232-260.
(2) Bockholt, H.; Udvarnoki, G.; Rohr, J.; Mocek, U.; Beale, J. M.; Floss, H.
G. J. Org. Chem. 1994, 59, 2064-2069.
Having confirmed the feasibility of using the amide group as a
precursor for benzopyrone synthesis, we set out to rationally design
this reactivity into an aromatic polyketide biosynthetic pathway.
We focused on the engineered biosynthesis of benzopyrone 5
(Scheme 1), which is an analogue of the HMG-CoA reductase
inhibitor pannorin.11 We reasoned that the benzopyrone ring
(Scheme 1) can be similarly generated from an amidated polyketide
intermediate 4, in which a nucleophilic phenol has been positioned
five atoms away from the amide carbonyl at C-19 to facilitate
pyrone formation. To satisfy these structural requirements, two
intramolecular aldol condensations must take place between C-9/
C-14 and C-7/C-16, in contrast to the C-7/C-12 regioselectivity
required for the synthesis of 15. We chose the TcmN cyclase from
the tetracenomycin pathway to control cyclization regiochemistry
because it has been shown to catalyze the desired cyclizations on
an acetate-primed decaketide.12 When OxyD and TcmN were
expressed in S. coelicolor with the oxy minimal PKS, the expected
metabolite WJ150 (5) was produced at a titer of 50 mg/L.
In summary, this work demonstrated that, in addition to amide-
and pyridone-containing compounds, a primary amide moiety can
also serve as an electrophilic center when positioned favorably,
readily yielding pyrone-containing polyketides. Benzopyrones are
(3) Terui, Y.; Chu, Y. W.; Li, J. Y.; Ando, T.; Yamamoto, H.; Kawamura,
Y.; Tomishima, Y.; Uchida, S.; Okazaki, T.; Munetomo, E.; Seki, T.;
Yamamoto, K.; Murakami, S.; Kawashima, A. Tetrahedron Lett. 2003,
44, 5427-5430.
(4) Li, A.; Piel, J. Chem. Biol. 2002, 9, 1017-1026.
(5) Zhang, W. J.; Ames, B. D.; Tsai, S. C.; Tang, Y. Appl. EnViron. Microbiol.
2006, 72, 2573-2580.
(6) Zhang, W.; Watanabe, K.; Wang, C. C.; Tang, Y. J. Nat. Prod. 2006, 69,
1633-1636.
(7) (a) Hiltunen, M.; Soderhall, K. Appl. EnViron. Microbiol. 1992, 58, 1043-
1045. (b) Dasenbrock, J.; Simpson, T. J. J. Chem. Soc., Chem Commun.
1987, 1235-1236.
(8) Koch, K.; Podlech, J.; Pfeiffer, E.; Metzler, M. J. Org. Chem. 2005, 70,
3275-3276.
(9) (a) Tanahashi, T.; Takenaka, Y.; Nagakura, N.; Hamada, N. Phytochemistry
2003, 62, 71-75. (b) Ihara, M.; Hirabayashi, A.; Taniguchi, N.; Fukumoto,
K. Heterocycles 1992, 33, 851-858.
(10) Kende, A. S.; Veits, J. E.; Lorah, D. P.; Ebetino, F. H. Tetrahedron Lett.
1984, 25, 2423-2426.
(11) Ogawa, H.; Hasumi, K.; Sakai, K.; Murakawa, S.; Endo, A. J. Antibiot.
1991, 44, 762-767.
(12) (a) McDaniel, R.; Hutchinson, C. R.; Khosla, C. J. Am. Chem. Soc. 1995,
117, 6805-6810. (b) Hutchinson, C. R. Chem. ReV. 1997, 97, 2525-
2536.
(13) (a) Fischer, C.; Lipata, F.; Rohr, J. J. Am. Chem. Soc. 2003, 125, 7818-
7819. (b) Xu, Z.; Jakobi, K.; Welzel, K.; Hertweck, C. Chem. Biol. 2005,
12, 579-588.
JA0736919
9
J. AM. CHEM. SOC. VOL. 129, NO. 30, 2007 9305