Formation of 1,3-cyclohexanediones and resorcinols catalyzed by a widely occuring ketosynthase

Article abstract of DOI:10.1002/anie.201210116

Overlooked, but widespread! A new class of ketosynthases (DarB) involved in the biosynthesis of 1,3-cyclohexanediones and dialkylresorcinols has been identified and characterized in detail. The presence of homologues in 89 different bacteria, including several pathogens, reveals that DarB as well as the corresponding natural products might be widespread, thus presenting a new but so far overlooked pathway to natural products. Copyright

Full text of DOI:10.1002/anie.201210116

.
Angewandte
Communications
DOI: 10.1002/anie.201210116
Biosynthesis
Formation of 1,3-Cyclohexanediones and Resorcinols
Catalyzed by a Widely Occuring Ketosynthase**
Sebastian W. Fuchs, Kenan A. J. Bozhꢀyꢀk, Darko Kresovic, Florian Grundmann,
Veronica Dill, Alexander O. Brachmann, Nicholas R. Waterfield, and
Helge B. Bode*
Angewandte
Chemie
4108
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4108 –4112

Angewandte
Chemie
Ketosynthases are essential for fatty acid and polyketide
biosynthesis, and several different classes of these enzymes
have been described. They can be part of large multifunc-
tional enzymes like the type I polyketide synthases (PKS) and
type I fatty acid synthases, or they exist as stand-alone
enzymes as in the type II fatty acid synthases, type II PKS,
or chalcone synthases (type III PKS).[1] In type II fatty acid
and polyketide biosynthesis, functionally different ketosyn-
thase classes are known, which either catalyze the first (KS III
or FabH) or the consecutive elongation steps (FabF and
FabB).[1,2] Despite their different functions, all ketosynthases
derivatives (9–11, Scheme 1b). Their structures were eluci-
dated by HR-ESI-MS (Table S4 in the Supporting Informa-
tion) and detailed 1D and 2D NMR analysis and (Table S5,
Figure S1). As NMR assignments were difficult owing to
keto–enol tautomerism, compounds 9–11 were methylated to
allow full NMR assignment (Table S6, Figure S1). Addition-
ally, the DAR compounds 12 and 13 were isolated from
Photorhabdus sp. PB 68.1, and their structures were con-
firmed by NMR spectroscopy (Table S5, Figure S1). Accord-
ing to the current hypothesis, DARs are formed by a DarA
(also named StlC) catalyzed condensation of two ACP-bound
b-keto-ACP-precursors (ACP = acyl carrier protein).[5,10]
However, this mechanism would not allow the formation of
CHD compounds like 9 and 10 as intermediates. Alterna-
tively, a consecutive Claisen condensation and Michael
addition of a b-keto- and an a,b-unsaturated-precursor
would allow the formation of a CHD compound as proposed
in the biosynthesis of 8,[14] subsequently leading to the
formation of DARs through oxidation.
À
are members of the thiolase protein family and catalyze C C
bond formation. However, recently the first member of the
KS III class has been described that is a functional malonyl-
transferase.[3]
Herein we report a novel and widespread class of
ketosynthases catalyzing the formation of 2,5-dialkylcyclo-
hexane-1,3-diones (CHDs) from two fatty acid derived
precursors. CHDs can be further oxidized to 2,5-dialkylre-
sorcinols (DARs) by an aromatase, the corresponding gene of
which is always encoded adjacent to the gene of the
ketosynthase. Since different biosynthesis gene clusters in
89 bacteria have been identified as using this novel biosyn-
thesis pathway, CHDs and DARs can be regarded as an
unexplored class of natural products. The only known
examples for DARs are resorstatin (1), DB-2073 (2),[4–6]
stemphol (3),[7] microcarbonin A (4),[8] resorcinin (5), and
isopropylstilbene (IPS) (6) and derivatives thereof from
entomopathogenic Photorhabdus strains (Scheme 1a).[9,10,11]
The flexirubins (7), found in several taxa of the gliding
bacteria, are DAR derivatives connected to a polyene acyl
chain showing antibacterial and anticancer activity.[12,13]
Hitherto, the only example of CHD natural products are
the chiloglottones (chiloglottone 1 (8)), which were identified
in orchids of the genus Chiloglottis, where they act as
pheromones to fool its pollinator, the wasp Neozeleboria
cryptoides.[14]
We decided to analyze the activity of DarAB by using
heterologously produced proteins from our model organism
and flexirubin producer Chitinophaga pinensis DSM 2588. As
previously described for darABC (darC encodes an acyl
When we analyzed the secondary metabolome of the g-
proteobacteria Photorhabdus sp. PB 68.1 and Photorhabdus
temperata subsp. thracensis, we identified three novel CHD
[*] Dipl.-Biol. S. W. Fuchs, Dipl.-Bioinf. K. A. J. Bozhꢀyꢀk,
Dipl.-Bioinf. D. Kresovic, F. Grundmann, M. Sc. V. Dill,
Dr. A. O. Brachmann, Prof. Dr. H. B. Bode
Merck Stiftungsprofessur fꢀr Molekulare Biotechnologie
Fachbereich Biowissenschaften
Max-von-Laue-Str. 9, 60438 Frankfurt am Main (Germany)
E-mail: h.bode@bio.uni-frankfurt.de
Dr. N. R. Waterfield
Department of Biology and Biochemistry
Building 3 South, Room 0.28, University of Bath
Claverton Down, Bath BA2 7AY (UK)
[**] This work was funded by the Deutsche Forschungsgemeinschaft
(DFG). We thank Harry R. Beller and Tim Schçner for providing
plasmids and enzymes, Peter Grꢀn for help with compound
purification, Imke Schmitt and Francesco Dal Grande for computing
time for phylogenetic analysis, and Michael Karas for MALDI
access.
Scheme 1. a) Structures of known DARs (1–7), a DAR derivative (7),
and the only example of a CHD natural product (8). b) CHD
derivatives (9–11) and DAR compounds (12, 13) identified during the
analysis described herein.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201210116.
Angew. Chem. Int. Ed. 2013, 52, 4108 –4112
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 4109

.
Angewandte
Communications
carrier protein) from Pseudomonas chlororaphis subsp.
aurantiaca,[5] the heterologous expression of darABC from
C. pinensis (cpin_6851, cpin_6850, and cpin_6845) in Escher-
ichia coli led to the production of the DAR compounds 14–17
(Scheme 2, Figure 1a), the structures of which were deter-
Scheme 2. Structures found after heterologous expression of darABC
from C. pinensis.
mined by HR-MS (Table S4) and NMR spectroscopy
(Table S7, Figure S2). On the other hand, the expression of
darBC did not result in the formation of DAR (Figure 1b) but
led to the identification of the CHD compounds 18–21
(Figure 1d), the structures of which were also determined by
HR-MS (Table S4) and NMR analysis (Table S8, Figure S2).
Regarding 18–21 to be precursors of 14–17, supports the DAR
biosynthesis via CHD formation. Thus, the ketosynthase
DarB catalyzes the cyclization of a b-keto-acyl precursor with
a a,b-unsaturated acyl precursor, thereby resulting in the
formation of a 4-carboxy-CHD, which is subsequently decar-
boxylated and oxidized. As expected, 4-carboxy derivatives of
all CHDs were detected after heterologous expression of
darBC (22–25; Figure 1 f); their structures were determined
from MS data only, as they were produced only in trace
amounts. No carboxy-CHDs were detected when darA was
additionally overexpressed (Figure 1e), thus indicating that
DarA converts carboxy-CHD or the carboxy-CHD-thioester
intermediates into the corresponding DARs. The CHDs
themselves (Figure 1d) might result from spontaneous thio-
ester hydrolysis and decarboxylation of carboxy-CHDs.
The DarAB homologues from Azoarcus BH72 (Fig-
ure S3), P. chlororaphis subsp. aurantiaca (Figure S4), and
Photorhabdus asymbiotica (Figure S5) were also analyzed
and led to the identification of additional carboxy-CHD,
CHD, and DAR compounds (Table S4). Thus, DarB homo-
logues from different organisms seem to exhibit different
specificities for the length of the cyclized fatty acid derived
precursors. Interestingly, the simultaneous overexpression of
darA from C. pinensis and darB from P. chlororaphis subsp.
aurantiaca also resulted in the conversion of a CHD (30; for
structure see Figure S4II) into the corresponding DAR (32;
Figure 1. HPLC/MS chromatograms of extracts from E. coli Bl21 (DE3)
Star expressing darABC (a, c, and e) and darBC (b, d, and f) from
C. pinensis. In a and b, c and d, and e and f extracted ion chromato-
grams (EIC) of the DAR compounds 14–17 (a), the CHD com-
pounds 18–21 (c), and the carboxylated CHD compounds 22–25
(g) are shown, respectively. Chromatogram f is enlarged 10-fold
relative to the intensities of b and d. Chromatograms a, c, and e are
enlarged 80-fold relative to the intensities of b and d.
Figure S4I), thus indicating that a cross-reaction between
DarA and DarB from different organisms is possible.
To confirm the proposed mechanism for CHD formation
by DarB and to show that the in vivo formation of CHD is
independent of additional E. coli components, an in vitro
assay with purified His6-DarB fusion protein was conducted.
The putative substrate b-ketopalmitoyl-coenzyme A (CoA)
was generated in situ as recently described.[15] Addition of
DarB and the second substrate butenoyl-CoA to b-ketopal-
mitoyl-CoA resulted in the production of 20 and 24 (Fig-
ure S6I, Table S6) as proposed (Scheme 3). We also tested
whether DarB can act as FabF by producing b-ketopalmitoyl-
CoA from myristoyl-CoA and malonylated ACP, but here no
production of 20 was observed (Figure S6I). Thus the
ketosynthase homologue DarB seems to have lost its ability
to catalyze chain-elongation reactions in favor of the ability to
catalyze the CHD formation from CoA-bound fatty acid
biosynthesis intermediates or degradation intermediates.
In vivo experiments (Figure 1) imply an aromatase activity
of DarA, which oxidizes CHD to the corresponding DAR. In
accordance with our finding, all described DarA homologues
4110
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4108 –4112

Angewandte
Chemie
A database search for DarAB led to the identification of
89 homologues in other bacteria (Table S10) counting only
strains encoding both DarA and DarB in close proximity. A
phylogenetic analysis of these DarB homologues together
with other ketosynthases revealed that DarB represents
a novel clade of ketosynthases (Figure 2, S12, S13). Compar-
ison of the phylogenetic trees of all identified DarA and DarB
homologues (Figure S14) suggests their coevolution.
Scheme 3. In vitro production of CHD 20 by DarB from C. pinensis
using an acyl-CoA oxidase and FadB as previously described for the
production of b-ketopalmitoyl-CoA.[15] The conversion of 20 or 24 into
16 was shown by heterologous expression of darABC in E. coli
(Figure 1).
Figure 2. Phylogenetic tree (PHYML) comprising different ketosyn-
thases. Details are listed in Table S11 and a zoomed version of the
DarB branch is shown in Figure S14. The arrow points to DarB from
the Gram-positive Nocardia brasiliensis ATCC 700358. The scale bar
indicates the degree of divergence as substitutions per sequence
position.
exhibit a weak similarity with flavin adenine dinucleotide
(FAD)-containing enzymes like flavodoxin and WrpA.[16,17]
Sequence alignments and homology modeling using the
crystal structure of FabH from Aquifex aeolicus VF5 and
Staphylococcus aureus resulted in good models for the DarB
homologues from C. pinensis DSM 2588, P. chlororaphis
subsp. aurantiaca, and P. asymbiotica (Figure S7) and allowed
the identification of the catalytic triad as C123-H297-N332 in
DarB from C. pinensis DSM 2588 (Figure S8)[18] as was
supported by a loss or decrease in CHD production after
production of C123A and N332A DarB variants in E. coli
(Figure S6II). As DarB shows a dead-end alkyl-chain tunnel
connected to the active site (Figure S9), a flip-flop catalysis
mechanism with an open and closed conformation as it has
been described for FabH from M. tuberculosis can be
assumed.[19] In contrast to the FabH catalysis mechanism,
the DarB model from C. pinensis DSM 2588 exhibits an
additional cavity, which allows access of a second CoA-bound
precursor to the active site (Figure S10). This observation is in
accordance with MALDI-MS data showing that DarB can
noncovalently bind two CoA-bound precursors (Figure S11).
Moreover, darA and darB are always part of operons
(Figure S15) as shown for example by a putative micro-
carbonin (4) biosynthesis gene cluster (Figure S15g) encoding
two predicted desaturase genes; a gene cluster in Methyl-
obacterium sp. 4–46 (Figure S15j) encoding two type I PKS;
the only gene cluster in a Gram-positive bacterium, Nocardia
brasiliensis ATCC 700358 (Figure S15f); and a putative
flexirubin biosynthesis gene cluster[20] in the human pathogen
Myroides odoratus (Figure S15k).
Considering the large number of bacterial genomes
encoding DarAB homologues, CHD or DAR compounds
appear to be a widespread natural product class, most of
which remain to be identified. Among the 89 bacterial strains
identified, 39 and five are human and animal pathogens,
respectively, including 15 Neisseria strains. Another nine
strains live in association with other organisms. Thus one
Angew. Chem. Int. Ed. 2013, 52, 4108 –4112
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4111

.
Angewandte
Communications
[8] D. Beresovsky, O. Hadas, A. Livne, A. Sukenik, A. Kaplan, S.
Carmeli, Isr. J. Chem. 2006, 46, 79 – 87.
[9] V. J. Paul, S. Frautschy, W. Fenical, K. H. Nealson, J. Chem. Ecol.
1981, 7, 589 – 597.
might speculate about a conserved function of these com-
pounds as virulence factors as shown for the antibiotic and
anticancer-active flexirubins,[13] the pluripotent isopropylstil-
bene (6) or as signaling compounds (e.g. 8[14]). Additionally,
resorstatin (1) has been described as a growth stimulator for
mammalian cells.[21] The simultaneous production of flexir-
ubin and resorstatin exhibiting antagonistic activities, might
indicate that DAR natural products play a role in the
modulation of eukaryotic host cells, an idea supported by
other known activities of these compounds.[11,13,22,23] Thus the
isolation of compounds corresponding to the identified
biosynthesis gene clusters must be a future goal in order to
assign their biological functions.
[10]
S. A. Joyce, A. O. Brachmann, I. Glazer, L. Lango, G. Schwꢀr,
D. J. Clarke, H. B. Bode, Angew. Chem. 2008, 120, 1968 – 1971;
Angew. Chem. Int. Ed. 2008, 47, 1942 – 1945.
[11] W. H. Richardson, T. M. Schmidt, K. H. Nealson, Appl. Environ.
Microbiol. 1988, 54, 1602 – 1605.
[12] H. Achenbach, W. Kohl, H. Reichenbach, Rev. Latinoam. Quim.
1978, 9, 111 – 124.
[13] N. Mojib, R. Philpott, J. P. Huang, M. Niederweis, A. K. Bej,
Antonie Van Leeuwenhoek 2010, 98, 531 – 540.
[14] S. Franke, F. Ibarra, C. M. Schulz, R. Twele, J. Poldy, R. A.
Barrow, R. Peakall, F. P. Schiestl, W. Francke, Proc. Natl. Acad.
Sci. USA 2009, 106, 8877 – 8882.
Received: December 18, 2012
[15] E. B. Goh, E. E. K. Baidoo, J. D. Keasling, H. R. Beller, Appl.
Environ. Microbiol. 2012, 78, 70 – 80.
Published online: February 19, 2013
[16] J. Sancho, Cell. Mol. Life Sci. 2006, 63, 855 – 864.
[17] J. Carey, J. Brynda, J. Wolfova, R. Grandori, T. Gustavsson, R.
Ettrich, I. K. Smatanova, Protein Sci. 2007, 16, 2301 – 2305.
[18] X. Qiu, C. A. Janson, A. K. Konstantinidis, S. Nwagwu, C.
Silverman, W. W. Smith, S. Khandekar, J. Lonsdale, S. S. Abdel-
Meguid, J. Biol. Chem. 1999, 274, 36465 – 36471.
[19] S. Sachdeva, F. N. Musayev, M. M. Alhamadsheh, J. N. Scarsdale,
H. T. Wright, K. A. Reynolds, Chem. Biol. 2008, 15, 402 – 412.
[20] M. J. McBride, G. Xie, E. C. Martens, A. Lapidus, B. Henrissat,
R. G. Rhodes, E. Goltsman, W. Wang, J. Xu, D. W. Hunnicutt,
A. M. Staroscik, T. R. Hoover, Y. Q. Cheng, J. L. Stein, Appl.
Environ. Microbiol. 2009, 75, 6864 – 6875.
Keywords: biosynthesis · cyclohexanedione · dialkylresorcinol ·
ketosynthase · natural products
.
[1] C. Hertweck, Angew. Chem. 2009, 121, 4782 – 4811; Angew.
Chem. Int. Ed. 2009, 48, 4688 – 4716.
[2] S. W. White, J. Zheng, Y. M. Zhang, C. O. Rock, Annu. Rev.
Biochem. 2005, 74, 791 – 831.
[3] T. Bretschneider, G. Zocher, M. Unger, K. Scherlach, T. Stehle,
C. Hertweck, Nat. Chem. Biol. 2012, 8, 154 – 161.
[4] S. Kato, K. Shindo, H. Kawai, M. Matsuoka, J. Mochizuki, J.
Antibiot. 1993, 46, 1024 – 1026.
[5] B. Nowak-Thompson, P. E. Hammer, D. S. Hill, J. Stafford, N.
Torkewitz, T. D. Gaffney, S. T. Lam, I. Molnar, J. M. Ligon, J.
Bacteriol. 2003, 185, 860 – 869.
[21] S. Imai, K. Fujioka, K. Furihata, R. Fudo, S. Yamanaka, H. Seto,
J. Antibiot. 1993, 46, 1319 – 1322.
[22] A. Pohanka, J. Levenfors, A. Broberg, J. Nat. Prod. 2006, 69,
654 – 657.
[6] N. Kanda, N. Ishizaki, N. Inoue, M. Oshima, A. Handa, T.
Kitahara, J. Antibiot. 1975, 28, 935 – 942.
[23] A. K. Bej, US 2011/030121A1 (PCT/US10/24823), 2011.
[7] F. H. Stodola, D. Weisleder, R. F. Vesonder, Phytochemistry
1973, 12, 1797 – 1798.
4112
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4108 –4112