Elucidating the biosynthetic pathway for vibralactone: A pancreatic lipase inhibitor with a fused bicyclic β-lactone

Article abstract of DOI:10.1002/anie.201208182

Not so simple: Evidence from ¹³C-labeling studies, metabolite profiling, and cell-free conversion established that the bicycle skeleton of vibralactone is derived from an aryl ring moiety and both shikimate and phenylalanine pathways may contri

Full text of DOI:10.1002/anie.201208182

.
Angewandte
Communications
DOI: 10.1002/anie.201208182
Natural Products
Elucidating the Biosynthetic Pathway for Vibralactone: A Pancreatic
Lipase Inhibitor with a Fused Bicyclic b-Lactone**
Pei-Ji Zhao,* Yan-Long Yang, Liangchen Du, Ji-Kai Liu, and Ying Zeng*
Vibralactone (1, Scheme 1a), a rare
fused bicyclic b-lactone from the
basidiomycete fungus Boreostereum
vibrans (syn. Stereum vibrans), was
found to inhibit pancreatic lipases
highly relevant to fat absorption.[1]
Inspired by its clinical potential for
the treatment of obesity, Zhou and
Snider developed an elegant 10-step
chemical route to the total synthesis
of (ꢀ)-vibralactone and extended
the work to (ꢁ)-vibralactone and
(ꢁ)-vibralactone C.[2] Recent inves-
tigations established that the
unusual fused b-lactone bicycle
system of vibralactone may account
for binding of both types of casein-
olytic peptidases vital for bacterial
virulence.[3] While vibralactone
seems structurally simple for its
small size, contradicting biosynthet-
Scheme 1. a) Vibralactone (1) and related compounds from the fungus B. vibrans. b) Labeling
ic origins, including sesquiterpenoid
patterns from feedings of 13C-labeled precursors. Bold lines indicate 13C-labeled isotopologue groups
and polyketide, of this bicyclic b-
with directly adjacent 13C atoms. Numbers represent the molar abundances of multiply 13C-labeled
lactone have been proposed.[4] Fur-
isotopologues. Arrows indicate 13C-isotopologue groups with two directly adjacent 13C atoms and
one 13C atom in the same molecule detected by long-range coupling in the 13C NMR spectra (see the
Supporting Information for details).
thermore, very little has been done
on the biosynthetic mechanisms for
natural products isolated from basi-
diomycete fungi in general.[5] We
therefore set out to investigate the
biosynthesis of vibralactone by 13C-labeling and metabolite
profiling studies, in addition to biochemical evidence. The
results unequivocally show that the bicyclic lactone core of
[*] Dr. P.-J. Zhao, Y.-L. Yang, Prof. Dr. J.-K. Liu, Prof. Y. Zeng
State Key Laboratory of Phytochemistry and Plant Resources in West
China, Kunming Institute of Botany, Chinese Academy of Sciences
Kunming 650201, Yunnan (China)
vibralactone is derived neither from a seemingly plausible
polyketide nor a sesquiterpenoid origin, but instead a shiki-
mate pathway. Additionally, VibPT is an aromatic prenyl-
transferase (PTase) potentially implicated in the biosynthesis
of vibralactone. Our findings provide the first insight into
a novel route for b-lactone formation in mushrooms.
E-mail: zhaopeiji@mail.kib.ac.cn
biochem@mail.kib.ac.cn
Homepage: http://english.kib.cas.cn/OurTearms/
Based on the six-carbon lactone skeleton and the dime-
thylallyl side chain of vibralactone (1), we initially speculated
that a polyketide pathway might be responsible for the
assembly of the lactone core with subsequent attachment of
a dimethylallyl moiety to eventually afford 1. To test this idea,
sodium [1-13C]acetate was fed, in the beginning, to the culture
of B. vibrans. We used a multiple-feeding procedure by adding
five aliquots (5 ꢀ 100 mg) of each precursor at days 12, 14, 16,
18, and 20 after inoculation. The culture was harvested after
an additional two days of incubation. The labeled products
(titer 10–110 mgLꢁ1) were isolated by the method described
in the Supporting Information (SI-1). The signal structures of
the 13C signals in the labeled 1 or 2 were analyzed according to
Prof. Dr. L. Du
Department of Chemistry, University of Nebraska-Lincoln
Lincoln NE 68588-0304 (USA)
Y.-L. Yang
University of Chinese Academy of Sciences
Beijing 100049 (China)
[**] This research was supported by grants from the National Natural
Science Foundation of China (31170061), and the National Basic
Research Program of China (973 program, 2009CB522300 and
2013CB127500). We are grateful to Wolfgang Eisenreich (Technical
University of Munich (Germany)), Zi-Hua Justin Jiang (Lakehead
University, Canada) and anonymous reviewers for their comments.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201208182.
2298
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 2298 –2302

Angewandte
Chemie
Werner et al.,[6] and further confirmed by two-dimensional
INADEQUATE experiments. The relative 13C abundance of
individual carbon atoms was calculated by comparison of
integrals of the 13C signal for the 13C-labeled and unlabeled
compounds. The absolute 13C abundance was determined by
quantitative NMR spectroscopy by analyzing 13C-coupled
taxonomically the same as Stereum genus.[1] Accordingly, we
hypothesized that a shikimate-derived aryl molecule might be
the precursor for developing the bicyclic skeleton of 1 despite
its apparently non-aryl structure. To test this hypothesis,
[U-13C]dl-phenylalanine (Phe) was fed and assimilated con-
sequently to produce 2b showing strong abundance (16–18%;
see SI-6 in the Supporting Information) and coupling at C1/
1
satellite signals in the H NMR spectra.[6]
To our surprise, no label from [1-13C]acetate was evident
by 13C NMR spectroscopy in any of the bicyclic lactone
carbon atoms in 1a (Scheme 1b). In contrast, a high level of
labeling was observed at the dimethylallyl carbon atoms C1’
and C3’ with a relative abundance of 14.9% and 23.9%,
respectively (see SI-2 in the Supporting Information). Feed-
ing of sodium [1,2-13C]acetate resulted in 2a with intact
incorporation of the acetate units at C1’/C2’ (J = 45.8 Hz),
C3’/C4’ (J = 41.8 Hz), and C3’/C5’ (J = 42.3 Hz) with 4.9–
6.5% abundances (see SI-3 in the Supporting Information),
thus leading directly to the evidence for the origin of the
dimethylallyl moiety through mevalonate as anticipated.
Likewise, no clear abundance of the bicyclic carbon atoms
could be discerned. This labeling pattern highlighted the
sharp distinction between the biosynthetic origins of each
moiety and ruled out the possibility of a polyketide pathway
which correlates to the bicyclic carbon backbone of 1.
Consequently, we hypothesized lysine to be the putative
precursor because lysine has six carbon atoms matching
perfectly with the carbon numbers of the bicyclic core
structure of 1. Nevertheless, no carbon atoms in 1 or its
congeners were clearly labeled on feeding with [1,2-13C]lysine.
We next administered the general precursor (uniformly
labeled) [U-13C]glucose to B. vibrans. This resulted in all the
carbon atoms being labeled in 1b and 2 with 3.8–11%
abundances (see SI-4 in the Supporting Information). As
shown in Scheme 1b, 13C3-isotopologue groups with three
contiguous 13C atoms were observed in 1b in addition to
13C4 isotopologues. Strong correlations were found between
C7 and C3, and C3 and C2 in the INADEQUATE NMR
spectra. This observation, in combination with the feeding
pattern from 13C acetate, suggested that molecules earlier
than acetyl-CoA in the glycolytic pathway must be the carbon
units from which the in vivo biosynthetic precursor pools are
formed for generating the bicyclic carbon skeleton of 1. The
presence of 13C3 isotopologues such as C2-C3-C7, and C4-C3-
C7 presumably indicates their common origin from a three-
carbon unit. This proposition was further supported by the
identical 13C3-isotopologue patterns in 1c from feedings of
[U-13C]glycerol, as shown in Scheme 1b and SI-5 in the
Supporting Information.
C2/C3/C4/C5/C6/C7 (J1-2 = 29.3 Hz,
42.5 Hz, J4-5 = 38.5 Hz, J5-1 = 29.3 Hz, J6-1 = 45.2 Hz, J7-3
J
2-3 = 41.9 Hz, J4-3
=
=
48.5 Hz), and very weak (ca. 2%) labeling at the dimethylallyl
carbon atoms. Acetyl-CoA originating from degradation of
Phe appears to be a rational source for the weak incorpo-
ration of the label at the dimethylallyl moiety. Significant
correlations of C6/C1/C2/C3(/C7)/C4/C5 were observed in
the INADEQUATE spectrum (SI-6), thus indicating a seven-
carbon unit originating from the benzenoid ring of Phe. This
observation was well corroborated by the feeding of [U-13C]4-
hydroxybenzoate, thus giving rise to the labeled product 2c
with distinctly high levels (43.4%; see SI-7a in the Supporting
Information) of intact incorporation of benzoate at C1/C2/C3/
C4/C5/C6/C7. Strong correlations exist between all seven
adjacent carbon atoms of the bicyclic skeleton, as shown in
the INADEQUATE spectrum of 2c (SI-7a). Using LC/MS
with selected ion monitoring, specific quasimolecular ions at
m/z 254 [M+7+Na]+ and m/z 264 [M+7+CH3OH+H]+ for
2c were distinctively recognized along with the normal 2 at
m/z 247 [M+Na]+ and m/z 257 [M+CH3OH+H]+, all corre-
sponding to the same retention time (see SI-7b in the
Supporting Information). The above labeling patterns con-
firmed that the bicyclic lactone core of 1 is derived from an
aryl ring that both Phe and shikimate routes may contribute
to. 4-Hydroxybenzoate can be considered as a candidate
precursor, which might be formed by two independent
pathways in this fungus: from chorismate by the chorismate
lyase reaction as normally known in bacteria and fungi, and
from Phe as validated by this research. The Phe route leading
to a C6C1 unit like 4-hydroxybenzoate occurs widely in plants
by way of Phe!cinnamate!4-coumarate, with subsequent
cleavage of two carbon atoms from the 4-coumarate side
chain. Whereas in a different way, the basidiomycete fungi
proceed as follows: Phe!cinnamate!benzoate!4-hydroxy-
benzoate, as confirmed from feedings of labeled l-Phe in
Bjerkandera adusta and Phanerochaete chrysosporium where
a ubiquitous lignin peroxidase can form 4-hydroxybenzoate
from benzoate.[8] Although the conversion of Phe into
tyrosine by phenylalanine hydroxylase has been well estab-
lished in animals, available evidence proved the absence of
this route in the basidiomycete fungi such as Stereum
sanguinolentun and Polyporus nidulans.[9] Nevertheless,
l-tyrosine can produce 4-hydroxybenzoate through 4-cou-
maroyl-CoA as verified in yeast.
The three-carbon unit of phosphoenolpyruvate is com-
bined with the four-carbon unit of erythrose 4-phosphate
from the pentose phosphate cycle to form shikimate, a funda-
mental precursor for aryl compounds biosynthesis. Although
the shikimate pathway is relatively less common in fungi than
in green plants, the basidiomycetes produced a number of
benzofurans which originate from shikimate pathway as
confirmed by 14C-labeling studies in Stereum subpileaturn.[7]
Besides, in our previous work a prenylated aryl compound (4,
Scheme 1a) was isolated as a minor metabolite from culture
broths of the fungus B. vibrans, which is the producer of 1 and
To probe the details hidden in the dramatic changes from
an aryl ring to the fused bicyclic b-lactone, we attempted
a time-course metabolite profiling approach. The culture
broths on days 10, 14, 18, and 22 (final) were analyzed by LC/
MS, and the results showed remarkable differences in product
profiles (Figure 1, and see SI-8 in the Supporting Informa-
tion). The broths on day 10 contained a very small amount of
1,5-secovibralactone (3) and a barely detectable amount of 1.
Angew. Chem. Int. Ed. 2013, 52, 2298 –2302
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 2299

.
Angewandte
Communications
Figure 1. LC/MS metabolite profiling of the time-course cultures. The
authentic standards (st) show quasimolecular ions [M+H]+ at m/z 209
for 3, [M+CH3OH+H]+ at m/z 241 for 1, and [M+CH3OH+H]+ at
m/z 257 for 2.
Figure 2. The cell-free enzymatic conversion of 3 (normal and labeled)
to 1 as confirmed by triple quadrupole LC/MS/MS.
The highest production of 3 was observed on days 14–18,
during which 1 started to form. From days 18 to 22, the
amount of compound 3 declined eventually to a barely
detectable level, while the amounts of 1 and 2 concomitantly
increased. The results suggest that 1 is likely to be derived by
cyclic rearrangement from 3, which may involve a novel ring
rearrangement mechanism. Interestingly, the new compound
5 (Scheme 1a and see SI-9a in the Supporting Information),
an isomer of 3, was identified as a minor metabolite from B.
vibrans in the feeding of [U-13C]4-hydroxybenzoate. As
confirmed by NMR analyses, 5 can afford 3, and vice versa.
From the same feeding, labeled [13C1-7]3 was also purified with
approximately 30% 13C abundance (see SI-9b in the Support-
ing Information). As a substrate, [13C1-7]3 would give the
labeled [13C1-7]1 which is highly specific to detection for
enzyme assays where formation of 1 from 3 can be
biochemically verified. We performed the in vitro enzymatic
conversion of 3 into 1 using a cell-free enzyme preparation
from mycelia of 20 days. When incubated with [13C1–7]3, the
crude enzyme eventually transformed 3 into 1 as a significant
product identified by LC/MS/MS. Compared to the boiled
enzyme control, the quasimolecular ions [M+7+K]+ at
m/z 254 for the labeled [13C1-7]1 and [M+K]+ at m/z 247 for
normal 1 were both observed exclusively in the enzymatic
product, all corresponding to the same retention time and
mass of the authentic standard 1 (normal) showing the
quasimolecular ion [M+K]+ at m/z 247 (Figure 2). As
expected, higher yield of the normal 1 (1.21e7, the base
peak) was observed in the enzymatic product since the
substrate is mainly composed of the normal 3 (SI-9b). The
collision induced MS/MS analysis of the parent ion [M+K]+ at
m/z 247 produced the dominant fragment ions at m/z 229 for
the loss of H2O, at m/z 215 for the loss of H2O and CH2, at
m/z 203 for the loss of CH2 and CHOH (probably at C7; see
SI-10 in the Supporting Information), and is in good agree-
ment with those of the standard 1. Specifically, the parent ion
[M+7+K]+ at m/z 254 showed + 7 fragmentation patterns
with dominant ions at m/z 236 (from 229 + 7) and m/z 222
(from 215 + 7), in addition to + 6 fragment ion at m/z 209
(from 203 + 6). The results additionally confirm the
enzymatic production of 1 from 3.
Next, we speculated that an aryl compound, such as 6
(Scheme 2), is most likely to be the early intermediate which
would give the seven-membered ring oxepin 7 by epoxidation
and rearrangement. To test this hypothesis, we synthesized 6
starting from 3,3-dimethylallyl bromide and 4-hydroxybenz-
aldehyde with subsequent reduction (see SI-11 in the
Supporting Information). Using 6 as a reference, we searched
through all the time-course cultures under LC/MS but failed
to detect the intermediate 6. The reason for this is probably
owing to the transient nature of 6 on the metabolic pathway of
Scheme 2. Proposed biosynthesis of vibralactone (1)
2300 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 2298 –2302

Angewandte
Chemie
this fungus. However, the in vivo transformation of 6a to the
new compound 8 (Scheme 2 and see SI-12 in the Supporting
Information) was observed within 24 hours after feeding the
14 day mycelia with 6a. Therefore, it is possible that an
oxidative expansion on the aryl ring can take place after
prenylation. To accommodate all of these observations, we
have proposed the following mechanism to account for this
unusual pathway as shown in Scheme 2. By a unique prenyl-
transferase, the dimethylallyl moiety from a common meval-
onate pathway might be attached to the aryl ring moiety that
both shikimate and phenylalanine routes may contribute to,
thus leading to the likely intermediate 6. The prenylation at
an aryl ring could be closely analogous to that proposed for
benzofuran biosynthesis in basidiomycete fungus Heterobasi-
dion occidentale and root cultures of Tagetes patula, respec-
tively.[10] Presumably 6 is required to undergo an epoxidation
to 7a and subsequent rearrangement to the oxepin inter-
mediate 7, which would rapidly tautomerize to the seven-
membered ring oxepinone 5 and then isomerize to 3. The
oxidation mechanism could potentially resemble that of the
final steps in the biosynthesis of aflatoxin B1 in Aspergillus,
where a ring expansion of a phenol to a lactone is mediated by
a cytochrome P450.[11] Eventually an intramolecular cycliza-
tion of 3 can afford 1, followed by an additional epoxidation
to 2. This unusual cyclization can be rationalized by formu-
lating a base-catalyzed deprotonation at C1 followed by
formation of an enolate which can then attack the ring enolate
to form the b-lactone system and protonation of the
neighboring carbon atom. It is also possible to assume an
initial H-abstraction from the sp3-CH ring atom, followed by
Cope rearrangement as proposed by Schwarzer et al.[12]
Alternatively, a single-electron oxidation of 3 at C1 may be
considered, followed by radical cyclization, which is strongly
favored by recent studies.[13]
Finally, we attempted to identify the prenyltransferase
(PTase) gene involved in vibralactone biosynthesis. From the
genomic DNA of B. vibrans we constructed a genomic library
followed by a probe hybridization for screening of candidate
fosmids. Random sequencing followed by bioinformatics
analysis enabled our discovery of a fosmid of about 30 kb
which harbors a putative aromatic PTase gene (vibPT) and
a putative cytochrome P450 monooxygenase (CYP) gene
(vibP450) nearby. The deduced amino acid sequence of vibPT
is 90% identical to its homologue ShPT recently found in the
genome of Stereum hirsutum (see SI-13 in the Supporting
Information), and shared 14–20% pairwise identity with the
verified aromatic PTases in ascomycete fungi such as indole
PTases (FgaPT2, SirD) and polycyclic PTases (VrtC),[14–16] and
grouped together with the putative aromatic PTases from
basidiomycete fungi (Figure 3A, shaded region). The crude
recombinant VibPT was shown to catalyze the formation of 6
from 4-hydroxybenzyl alcohol and dimethylallyl diphosphate
(Figure 3B and see SI-13). Furthermore, based on the RNA-
Seq the vibPT transcripts were significantly accumulated on
days 14 and 20, about 4–5-fold greater than that of day 7. The
relatively more abundant levels of vibPT expression on days
14 and 20 correlated well with the remarkably higher
production of 3, 1, and 2 during the same time (Figure 1).
Thus, VibPT can be considered as a potential enzyme to
Figure 3. A) Phylogenetic connection of VibPT and other fungal aro-
matic PTases. The tree scale was modified to adapt for the chemical
structures (see SI-13 in the Supporting Information). Shown in parallel
are the corresponding substrates which each enzyme prefers. B) HPLC
analysis of products formed by the crude recombinant VibPT and the
empty pET control.
catalyze the transfer of a dimethylallyl group to the aryl ring
moiety, and most likely dedicated to the vibralactone
biosynthesis.
In summary, we have established the biosynthetic origin of
vibralactone and obtained evidence for its biosynthetic
pathway which includes several very interesting reactions
which may involve unusual enzymes. The bicyclic lactone core
of vibralactone could be attributed to an intramolecular
cyclization of the seven-membered ring intermediate, which is
likely to arise from an oxidative expansion of the aryl ring
moiety that is derived from phenylalanine and shikimate
pathways. Vibralactone has a unique chemical structure and
potent biological activity, and is isolated from a basidiomycete
fungus. All of these attributes make vibralactone a worthy
target for biosynthetic investigation. The establishment of its
biosynthetic pathway will be valuable for future elucidation of
the molecular mechanism for its biosynthesis.
Received: October 11, 2012
Revised: November 22, 2012
Published online: January 21, 2013
Keywords: biosynthesis · isotopic labeling · lactones ·
.
mass spectrometry · natural products
[1] a) D.-Z. Liu, F. Wang, T.-G. Liao, J.-G. Tang, W. Steglich, H.-J.
Zhu, J.-K. Liu, Org. Lett. 2006, 8, 5749 – 5752; b) M.-Y. Jiang, F.
Wang, X.-L. Yang, L.-Z. Fang, Z.-J. Dong, H.-J. Zhu, J.-K. Liu,
Chem. Pharm. Bull. 2008, 56, 1286 – 1288.
[2] a) Q. Zhou, B. B. Snider, Org. Lett. 2008, 10, 1401 – 1404; b) Q.
Zhou, B. B. Snider, J. Org. Chem. 2008, 73, 8049 – 8056.
Angew. Chem. Int. Ed. 2013, 52, 2298 –2302
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2301

.
Angewandte
Communications
[3] E. Zeiler, N. Braun, T. Bottcher, A. Kastenmuller, S. Weinkauf,
S. A. Sieber, Angew. Chem. 2011, 123, 11193 – 11197; Angew.
Chem. Int. Ed. 2011, 50, 11001 – 11004.
[4] a) R. A. Hill, A. Sutherland, Nat. Prod. Rep. 2007, 24, 14 – 17;
b) B. M. Fraga, Nat. Prod. Rep. 2009, 26, 1125 – 1155; c) A.
Schꢁffler, T. Anke in The Mycota XV, Physiology and Genetics,
Vol. 15 (Eds.: T. Anke, D. Weber), Springer, Heidelberg, 2009,
pp. 209 – 231.
[5] a) P. Spiteller, M. Spiteller, W. Steglich, Angew. Chem. 2003, 115,
2971 – 2974; Angew. Chem. Int. Ed. 2003, 42, 2864 – 2867; b) M.
Lang, A. Muhlbauer, E. Jagers, W. Steglich, Eur. J. Org. Chem.
2008, 3544 – 3551.
[6] a) I. Werner, A. Bacher, W. Eisenreich, J. Biol. Chem. 1997, 272,
25474 – 25482; b) D. Arigoni, S. Sagner, C. Latzel, W. Eisenreich,
A. Bacher, M. H. Zenk, Proc. Natl. Acad. Sci. USA 1997, 94,
10600 – 10605; c) D. Eichinger, A. Bacher, M. H. Zenk, W.
Eisenreich, J. Am. Chem. Soc. 1999, 121, 7469 – 7475.
[7] a) J. D. BuꢂLocK, A. T. Hudson, B. Kaye, Chem. Commun. 1967,
814 – 815; b) J. D. BuꢂLock, B. Kaye, A. T. Hudson, Phytochem-
istry 1971, 10, 1037 – 1046.
K. M. C. Evans, T. K. Kirk, K. E. Hammel, Appl. Environ.
Microbiol. 1994, 60, 709 – 714.
[9] P. Chandra, L. C. Vining, Can. J. Microbiol. 1968, 14, 573 – 578.
[10] a) L. Margl, C. Ettenhuber, I. Gyurjꢅn, M. H. Zenk, A. Bacher,
W. Eisenreich, Phytochemistry 2005, 66, 887 – 899; b) D. Hans-
son, A. Menkis, ꢆ. Olson, J. Stenlid, A. Broberg, M. Karlsson,
Phytochemistry 2012, 84, 31 – 39.
[11] D. W. Udwary, L. K. Casillas, C. A. Townsend, J. Am. Chem. Soc.
2002, 124, 5294 – 5303.
[12] D. D. Schwarzer, P. J. Gritsch, T. Gaich, Angew. Chem. 2012, 124,
11682 – 11684; Angew. Chem. Int. Ed. 2012, 51, 11514 – 11516.
[13] a) S. M. Kroner, M. P. DeMatteo, C. M. Hadad, B. K. Carpenter,
J. Am. Chem. Soc. 2005, 127, 7466 – 7473; b) J. Yin, C. Wang, L.
Kong, S. Cai, S. Gao, Angew. Chem. 2012, 124, 7906 – 7909;
Angew. Chem. Int. Ed. 2012, 51, 7786 – 7789.
[14] a) I. A. Unsçld, S.-M. Li, Microbiology 2005, 151, 1499 – 1505;
b) N. Steffan, I. A. Unsçld, S.-M. Li, ChemBioChem 2007, 8,
1298 – 1307.
[15] a) A. Kremer, S.-M. Li, Microbiology 2010, 156, 278 – 286; b) S.-
M. Li, Phytochemistry 2009, 70, 1746 – 1757.
[8] a) C. Lapadatescu, C. Giniꢃs, J.-L. Le Quꢄrꢄ, P. Bonnarme, Appl.
Environ. Microbiol. 2000, 66, 1517 – 1522; b) K. A. Jensen,
[16] Y.-H. Chooi, P. Wang, J. Fang, Y. Li, K. Wu, P. Wang, Y. Tang, J.
Am. Chem. Soc. 2012, 134, 9428 – 9437.
2302 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 2298 –2302