Genomics-driven discovery of taiwachelin, a lipopeptide siderophore from Cupriavidus taiwanensis

Article abstract of DOI:10.1039/c2ob26296g

A genome mining study led to the identification of a previously unrecognised siderophore biosynthesis gene cluster in the nitrogen-fixing bacterium Cupriavidus taiwanensis LMG19424. Based upon predicted structural residues, a convenient strategy for an NMR-assisted isolation of the associated metabolite was designed. The structure of the purified siderophore, taiwachelin, was fully characterized by spectroscopic methods and chemical derivatisation.

Full text of DOI:10.1039/c2ob26296g

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
www.rsc.org/obc
Volume 10 | Number 47 | 21 December 2012 | Pages 9297–9508
ISSN 1477-0520
PAPER
Martin F. Kreutzer and Markus Nett
Genomics-driven discovery of taiwachelin, a lipopeptide siderophore
from Cupriavidus taiwanensis
1477-0520(2012)10:47;1-4

Dynamic Article Links
Organic &
View Article Online
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 9338
www.rsc.org/obc
PAPER
Genomics-driven discovery of taiwachelin, a lipopeptide siderophore from
Cupriavidus taiwanensis†
Martin F. Kreutzer and Markus Nett*
Received 6th July 2012, Accepted 30th August 2012
DOI: 10.1039/c2ob26296g
A genome mining study led to the identification of a previously unrecognised siderophore biosynthesis
gene cluster in the nitrogen-fixing bacterium Cupriavidus taiwanensis LMG19424. Based upon predicted
structural residues, a convenient strategy for an NMR-assisted isolation of the associated metabolite was
designed. The structure of the purified siderophore, taiwachelin, was fully characterized by spectroscopic
methods and chemical derivatisation.
structural and physicochemical features that are typically associ-
ated with siderophores from oceanic bacteria, but only rarely
Introduction
The development of cost-effective high-throughput sequencing
technology and the resulting surge in whole genome sequences
has promoted a paradigm shift in natural product research in the
past few years. Rather than relying on bioactivity-guided screen-
ing for the discovery of useful molecules, many research groups
nowadays exploit the information that derives from genomic data
to identify new biosynthetic pathways and their associated
chemical compounds.1 This strategy, which is commonly
referred to as genome mining, is particularly useful in the case
of siderophores, small molecules that are secreted by micro-
organisms and plants to mobilize ferric iron from the environ-
ment under conditions when this transient metal becomes
scarce.2 Several siderophores exhibit a peptidic structure, which
is assembled through the concerted action of nonribosomal
peptide synthetases (NRPSs) or NRPS-independent siderophore
synthetases (NIS).3 The multifunctional NRPS complexes typi-
cally inhere a co-linear logic, in which the arrangement of cataly-
tic domains determines the sequence of enzymatic reactions.3a
Through analyses of their substrate-specifying domains, it is
often possible to predict the amino acid building blocks that are
utilised during the assembly process.4 This knowledge, together
with the iron-responsive production of siderophores, facilitates
the isolation and structure elucidation of new members of this
compound class.5
found in other microbial groups, namely a lipopeptide backbone
and photoreactivity.7 We were hence interested in whether the
production of cupriachelin is specific to C. necator H16, or
whether other taxonomically related strains also depend on such
compounds for iron sequestration. To clarify this, we analysed
available genomic sequences from soil-derived and plant-associ-
ated members of this bacterial group for the presence of sidero-
phore biosynthesis pathways that bear genetic hallmarks of fatty
acid incorporation. In the case of cupriachelin, the attachment of
the fatty acid side-chain was proposed to be catalysed by the
starter condensation domain of an NRPS.6 Alternative mechan-
isms, which are also conceivable for lipidation reactions, involve
the combined action of a fatty acyl-AMP ligase and an acyl
carrier protein (ACP) or require a single ACP alone.8 By inte-
grating the search for fatty acid-loading enzymes in the
sequence-based screening for novel siderophore biosynthesis
pathways, it should hence be possible to identify loci that code
for the production of lipopeptide iron(III) ligands.
Here, we report the successful application of this genome
mining strategy resulting in the discovery of the taiwachelin bio-
synthesis gene cluster in the nitrogen-fixing bacterium Cupriavi-
dus taiwanensis LMG19424. The product of the identified
pathway represents a previously unknown lipopeptide sidero-
phore, which was isolated and physicochemically characterized
as part of this study.
We recently reported the genomics-inspired discovery of an
NRPS-derived siderophore from the freshwater bacterium
Cupriavidus necator H16 (syn. Ralstonia eutropha H16).6 The
isolated compound, cupriachelin, is unusual in that it exhibits
Results and discussion
In silico screening for pathways to lipopeptide siderophores
Leibniz Institute for Natural Product Research and Infection Biology,
HKI, Beutenbergstr. 11a, D-07745 Jena, Germany.
E-mail: Markus.Nett@hki-jena.de; Fax: +49 3641 5320811;
Tel: +49 3641 5321297
†Electronic supplementary information (ESI) available. See DOI:
10.1039/c2ob26296g
For the bioinformatic analysis, we selected 20 environmental
strains of the family Burkholderiaceae whose genomes had been
fully assembled, annotated and deposited in GenBank. The
chosen strains represent the genera Burkholderia, Cupriavidus,
9338 | Org. Biomol. Chem., 2012, 10, 9338–9343
This journal is © The Royal Society of Chemistry 2012

View Article Online
Ralstonia and Polynucleobacter.9 Their genomic sequences were
initially screened for the presence of NRPS and NIS encoding
genes, using biochemically characterised representatives as
driver sequences in BLASTP analyses according to a previously
described method.10 To identify putative siderophore loci, the
genomic regions surrounding the primary hits were subsequently
analysed for the presence of open reading frames involved in
siderophore export and uptake, as well as hydroxamate, salicy-
late and catecholate formation.
locus that met our search criteria for the biosynthesis of a lipo-
peptide siderophore. The corresponding gene cluster was
detected on chromosome 2 of the rhizobium C. taiwanensis
LMG19424,13 and includes 18 ORFs between RALTA_B1222
and RALTA_B1239 covering 41.9 kb of contiguous DNA; the
genes of the cluster, which is in the following referred to as
taiwachelin or tai cluster, have been renamed in this study from
taiA to taiR (Table S2†).
Except for the endosymbiotic Polynucleobacter species,
which possess comparatively small genomes, all investigated
strains were found to harbor siderophore biosynthesis genes that
were distributed over one or two independent loci (Table S1†).
While the analysed Burkholderia genomes feature only NRPS-
dependent siderophore pathways, NIS genes appear to be quite
common in the genera Ralstonia and Cupriavidus. In many
cases, the identified loci could be associated with the production
of known siderophores based upon genetic precedence. The orni-
bactin assembly line appears thus to be one of the most prevalent
siderophore biosynthetic machineries among environmental
Burkholderia strains, which is consistent with previous studies
on human pathogenic Burkholderia isolates.11 In contrast, gene
clusters for the biosynthesis of micacocidin, yersiniabactin or sta-
phyloferrin B are commonly encountered in Ralstonia solana-
cearum genomes.12 Among the gene clusters, which could not
be correlated with a known product, we identified one candidate
Architecture of the tai gene cluster and proposed biosynthesis
The gene cluster comprises four successive NRPS genes
(taiDEFG), which are proposed to assemble the peptidic back-
bone of the siderophore (Fig. 1). The MbtH-like protein TaiB,
the type-II thioesterase TaiC and the phosphopantetheinyl trans-
ferase TaiQ have possible roles in activating and tuning the
NRPS biosynthetic machinery,14 whereas the lipase TaiN and the
monooxygenase TaiP are likely involved in supplying a fatty
acid starter unit and the amino acid building block Nδ-hydroxy-
Orn, respectively. The cyclase TaiA and the acyltransferase TaiO
can be envisioned to catalyze tailoring reactions. Five genes
within the cluster (taiH, taiI, taiK, taiM, and taiR) are associated
with siderophore-mediated iron transport and turnover.
Sequence alignments suggest a functional relatedness of the
N-terminal AMP-binding domain of TaiD to acyl-AMP ligases
that prime lipid starter units.15 TaiD is therefore proposed to
Fig. 1 Organisation of the tai biosynthesis gene cluster (top). Molecular processing line deduced from taiD–G (bottom). Domain notation: FAAL,
fatty acyl-AMP ligase; ACP, acyl carrier protein; C, condensation; TauD, hydroxylase; A, adenylation; PCP, peptidyl carrier protein; E, epimerase; TE,
thioesterase.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 9338–9343 | 9339

View Article Online
initiate the biosynthesis by loading an unspecified fatty acid onto
the NRPS assembly line. The consecutive incorporation of
monomeric amino acids is then carried out by the six NRPS
extension modules that are encoded by taiE, taiF and taiG. The
respective building blocks were predicted from the specificity-
conferring code of adenylation domains,4 and their configuration
in the final natural product was deduced from the presence or
absence of epimerisation domains in the NRPS modules.16
According to these analyses, the fully assembled siderophore
should feature L-Asp, D-allo-Thr, L-Asp, D-Nδ-hydroxy-Orn,
L-Ser and L-Nδ-hydroxy-Orn as amino acid constituents. The
C-terminal TauD domain in TaiD may account for the hydroxy-
lation of either or both predicted Asp residues. TaiO is closely
related to CucL, an enzyme implicated in the acylation of
Nδ-hydroxy-Orn with a 3-hydroxybutanoyl (Hbu) moiety.6 We
assumed a similar reaction in taiwachelin biosynthesis, as the
resulting hydroxamate function could serve as a bidentate ligand
for the coordination of ferric iron. It was unclear, however,
whether either or both Nδ-hydroxy-Orn moieties are converted
by TaiO. The termination module of TaiG features a thioesterase
domain and should thus catalyse the hydrolytic offload of the
acyl-S-enzyme intermediate from the assembly line.
chromatographic comparison of the two extracts, it was evident
that the observed CAS activity in the former was likely due to
one peak (Fig. 2). To obtain structural evidence that the ident-
ified peak may actually represent the sought molecule, the crude
extract from the iron-deficient culture was subjected to NMR
analyses prior to starting the isolation process. The assignment
of amino acid-related proton resonances is generally difficult in
extracts of heterotrophically grown microorganisms, because of
signal overlapping with media components. We thus decided to
look for signals that would indicate the predicted incorporation
of the Hbu residue. Previous studies on Hbu-featuring sidero-
phores had revealed that the methylene protons of acylated Hbu
exhibit resonances in the 1H NMR spectrum, which can be
easily distinguished from those of media-derived amino acids
within the same chemical shift range based upon their distinctive
double doublet splitting pattern.6,21 Furthermore, we evaluated
1
the presence of hydroxamate groups by a H,15N HMBC exper-
iment, as the corresponding correlations are quite specific and
only rarely detected in extracts from bacterial cultures. Both
NMR-based analyses supported the assumption that the extract
contained the product of the tai gene cluster. While the 1H NMR
spectrum showed discrete dd signals between 2.40 and
1
A search for the proposed peptide sequence in a natural
product database17 suggested that the product from the tai gene
cluster is likely to be novel, thereby minimising the chance to
retrieve a known compound in a subsequent fermentation study.
3.00 ppm, we also detected H,15N HMBC-derived connectiv-
ities that could be ascribed to long-range couplings with hydro-
xamate nitrogens in addition to typical amine and amide proton
correlations (Fig. S11†). HPLC clean-up was then guided by
CAS activity and NMR analyses, respectively, as the UV profile
of the identified peak lacked a conspicuous chromophore. This
approach yielded 20 mg of taiwachelin, which was isolated as a
white crystalline powder.
Genomics-guided isolation of taiwachelin
Select changes in media composition are known to increase or
even trigger natural product biosynthesis.18 In order to induce
the production of the predicted siderophore, we cultivated C.
taiwanensis LMG19424 in an iron-deficient H-3 minimal
medium. This approach appeared highly promising, since the tai
gene cluster represents the only putative siderophore biosynthesis
locus that was identified on the C. taiwanensis genome during
our in silico screening. To recover the siderophore from the fer-
mentation broth, we resorted to the adsorber resin XAD-2, which
has proven useful for the adsorption of amphiphilic sidero-
phores.19 The extract that was eluted from the adsorber resin
gave a strong response in the chrome azurol S (CAS) assay,20
confirming the presence of iron(III)-chelating molecules. In con-
trast, the extract from an iron-saturated C. taiwanensis culture
did not decolourise the iron dye complex. From the
Structure elucidation of taiwachelin
The FT-IR spectrum of taiwachelin (1, Fig. 3) indicated the pres-
ence of hydroxyl and amide functions, as expected from the
bioinformatic prediction. High resolution ESI-MS yielded m/z
963.4879 for the [M + H]+ ion, which is consistent with a mo-
lecular formula of C41H70O18N8 and corresponds to 11 degrees
of unsaturation. The double bond equivalents were assigned to
10 carbonyl moieties and one ring structure based upon an
inspection of the resonances in the 13C spectrum. H,1H COSY
1
data enabled an elucidation of the amino acid constituents of tai-
wachelin by resolving the individual spin systems of β-hydroxy-
Asp, Thr, Asp, Ser as well as two discrete Nδ-hydroxy-Orn resi-
dues. One of the latter two had to be cyclic, as its δ-methylene
protons showed strong long-range interactions in the HMBC
spectrum with the associated carbonyl carbon of the same spin
system. The second Nδ-hydroxy-Orn residue was found to be
Fig. 2 Chromatographic profiles of extracts from C. taiwanensis
LMG19424 grown in an H-3 medium under iron-deficient conditions
(A) and in the presence of 190 μM Fe(NH4) citrate (B).
Fig. 3 Structure of taiwachelin (1).
9340 | Org. Biomol. Chem., 2012, 10, 9338–9343
This journal is © The Royal Society of Chemistry 2012

View Article Online
N-acylated with a 3-hydroxybutanoyl moiety. The combined
results from the 1H,1H COSY, DEPT135, and HSQC spectra
eventually revealed a dodecanoyl moiety, thus concluding the
assignment of the 41 carbon resonances of taiwachelin. The
sequence of the six amino acids and the fatty acid was initially
established from HMBC correlations between the amide protons
structure of the metabolite was predicted with sufficient accuracy
from the molecular processing line to enable an NMR-assisted
isolation. Furthermore, the knowledge about the domain archi-
tecture of the taiwachelin biosynthesis enzymes expedited the
determination of the absolute configuration, underlining the great
potential of genome mining techniques.1 Taiwachelin shares
many structural features and physicochemical properties with
siderophores from oceanic bacteria.7 This resemblance is particu-
larly evident in comparison to loihichelin D, a photoreactive
lipopeptide siderophore from a marine Halomonas sp., which
also features a dodecanoate moiety and a terminal cyclic Nδ-
hydroxy-Orn residue.19 The production of photoreactive, lipo-
peptide iron(III) ligands appears thus not to be a characteristic of
the previously investigated C. necator strain H16, but is probably
more common among taxonomically related bacteria.
Further investigation on the role of taiwachelin’s photoreactivity
as well as cellular uptake experiments are in progress in our
laboratory.
2
and their J-coupled carbonyl carbons. MALDI-TOF/TOF frag-
mentation then proved the deduced planar structure (Fig. S1†).
To resolve the configuration of the single amino acids in taiwa-
chelin, we applied Marfey’s method following an acid-catalysed
hydrolysis of the native natural product.22 This study unequivo-
cally confirmed the exclusive presence of the single enantiomers
L-threo-β-hydroxy-Asp, D-allo-Thr, L-Asp and L-Ser in the acid
hydrolysate. In the case of Nδ-hydroxy-Orn, both the D- and the
L-form were identified (Table S4†). Although Marfey’s analysis
provided no information on the position of the two Nδ-hydroxy-
Orn enantiomers in the natural product, it is reasonable to
assume the sequence depicted in Fig. 3, which is fully consistent
with the arrangement of E domains in the taiwachelin assembly
line. The 3-hydroxybutanoyl (Hbu) moiety was determined to be
in the S configuration by chiral GC analysis following an esta-
blished protocol.6
Experimental
General experimental procedures
LC-MS experiments for metabolic profiling and Marfey’s analy-
sis were conducted on an Agilent 1100 series HPLC-DAD
system coupled with an MSD trap (Agilent) operating in alternat-
ing ionization mode and an Antek 8060 HPLC-CLN-detector
(Antek Instruments GmbH) using a C8 column (Zorbax Eclipse
XDB C8, 150 × 4.6 mm, 5 μm; Agilent). A linear gradient of
methanol in water + 0.1% trifluoroacetic acid (10% → 90%
methanol within 15 min; flow rate 1 mL min−1) was used for
metabolic profiling. Analytical HPLC for Marfey’s analysis of
serine was conducted on a Shimadzu UFLC liquid chromato-
graphy system equipped with a Nucleosil 100 C18 column (250
× 4.6 mm, 5 μm; CS Chromatographie Service). The separation
was accomplished by using a linear gradient from solvent A (1%
methanol, 5% acetonitrile in 10 mM ammonium formate) to
solvent B (1% methanol, 60% acetonitrile in 10 mM ammonium
formate) over 40 min with a flow rate of 1 mL min−1 and wave-
length monitoring at 365 nm.23 High resolution mass determi-
nation was carried out using an Exactive Mass Spectrometer
(Thermo-Scientific). One- and two-dimensional MALDI-TOF
MS data using post-source decay were acquired on a Bruker
Ultraflex spectrometer (Bruker Daltonics). NMR spectra were
recorded at 300 K on a Bruker Avance III 500 MHz spectrometer
with dimethylsulfoxide-d6 as a solvent and internal standard.
1H,15N HMBC spectra were referenced externally to urea. IR
spectra were obtained with a benchtop FT/IR 4100 spectrometer
(JASCO).
Complexing properties and photoreactivity of taiwachelin
It is obvious that the structure of taiwachelin features three biden-
tate coordination sites for iron(III). The six oxygen ligands come
from the two hydroxamate functions and the β-hydroxy-Asp
residue, allowing the formation of an octahedral complex. MS
studies were carried out to test the coordination ability of taiwa-
chelin for divalent and trivalent metal ions, respectively. Similar
to cupriachelin, the siderophore from C. taiwanensis selectively
chelated Fe3+, but did not coordinate Mn2+, Co2+, Ni2+, Cu2+,
and Zn2+. The stoichiometry of the complex formed between
iron(III) and taiwachelin was 1 : 1 (Fig. S3†).
The structural features of taiwachelin, particularly the presence
of a β-hydroxy-Asp moiety, suggested a susceptibility to UV
light when complexed to iron(III). Marine lipopeptide sidero-
phores and also cupriachelin were shown to undergo photoreduc-
tive dissociation under these conditions, which is accompanied
by the release of ferrous iron.6,7 To explore whether taiwachelin
is likewise prone to photochemical reactions, we monitored the
shifting of the iron(II) level in an aqueous solution of its ferric-
ion complex after exposure to sunlight. For quantification, Fe2+
was trapped with the specific ligand bathophenanthroline disul-
fonate (BPDS) and the absorbance of the developing coordi-
nation complex was measured at 535 nm. Control reactions that
were kept in the dark remained at very low absorption values of
0.003 0.002 during the entire assay, whereas the absorbance of
sunlight-treated samples increased from 0.001
0.001 to
0.073 0.006 in the same period. MS analyses confirmed the
loss of the intact iron(III)–taiwachelin complex in the latter.
Bioinformatic studies
Genomic sequences of bacteria were retrieved from the National
Center for Biotechnology Information (NCBI) database (http://
www.ncbi.nlm.nih.gov). These sequences were screened for
putative pathways to lipopeptide siderophores using homology-
based alignments against a handmade library of conserved bio-
synthesis genes.10 Initial hits were grouped by physical
Conclusions
We successfully identified a novel iron-chelating natural product
from the nitrogen-fixing bacterium C. taiwanensis LMG19424
by a combined genomic and analytical screening approach. The
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 9338–9343 | 9341

View Article Online
proximity and mapped onto the respective genome using Vector
NTI (Invitrogen).
(C-28), 164.9 (C-1), 169.4 (C-26), 169.9 (C-7), 170.0 (C-22),
170.3 (C-18), 171.4 (C-9), 171.4 (C-14), 171.9 (C-21), 172.7
(C-30), 173.0 (C-29). HR-ESI-MS: m/z 963.4879 [M + H]+
(calcd for C41H71O18N8, 963.4881).
Growth conditions
For the isolation of taiwachelin, C. taiwanensis LMG19424 was
grown in an H-3 mineral medium lacking ammonium ferric
citrate: 1 g L−1 aspartic acid, 2.3 g L−1 KH2PO4, 2.57 g L−1
Na2HPO4, 1 g L−1 NH4Cl, 0.5 g L−1 MgSO4·7H2O, 0.5 g L−1
NaHCO3, 0.01 g L−1 CaCl2·2H2O, and 5 mL L−1 SL-6 trace
element solution.7 The strain was shaken (160 rpm) at 30 °C for
4 days. The influence of Fe3+ on taiwachelin production was
tested in the same medium containing 190 μM ammonium ferric
citrate.
Assignment of the absolute configuration
The configurations of the amino acid constituents of taiwachelin
were determined following acid hydrolysis and derivatisation
with Marfey’s reagent (1-fluoro-2,4-dinitrophenyl-5-L-alanine
amide, L-FDAA, Sigma-Aldrich) by coelution experiments with
L-FDAA-derivatised amino acid standards. To this end, 500 μg
purified taiwachelin was dissolved in 400 μL concentrated HI
and heated at 110 °C for 3 h. The solution was lyophilised, and
the dried hydrolysate was resuspended in 10 μL of water and
20 μL of 1 M NaHCO3. Derivatisation was carried out with
170 μL of 1% L-FDAA in acetone at 37 °C for 1 h. The products
were lyophilised and prepared for LC-MS analysis by dissolving
in 1 mL of 50% acetonitrile. Standards for co-chromatography
were prepared by reacting 50 μL of 50 mM aqueous amino acid
solution with 20 μL of 1 M NaHCO3 and 100 μL of 1%
L-FDAA in acetone at 37 °C for 1 h. The lyophilised products
were then dissolved in 1 mL of 50% acetonitrile. LC-MS analy-
sis was performed under the aforementioned conditions with a
detection wavelength of 365 nm. The configuration of Hbu was
determined after hydrolysing 500 μg taiwachelin with 1 mL of
0.3 N HCl. The reaction was carried out at room temperature for
4 days after which excess reagent was removed under vacuum.
The hydrolysate was resuspended in water and purified over a
Sep-Pak RP-18 cartridge. Following the evaporation of the
eluent, 100 μL of trifluoroacetic anhydride–methylene chloride
(1 : 1, v/v) was added, and the mixture heated at 100 °C for
10 min. Excess reagents were removed with a stream of argon.
Methylene chloride, 100 μL, was added, and the resulting tri-
fluoroacetyl ester analysed using a Trace GC Ultra (Thermo)
coupled in parallel to an FID detector and a Polaris-Q ion trap
mass detector (Thermo). Separation was established on a chiral
silica column (25 m × 0.25 mm ID) coated with Chir-D-Val
(Varian) using a continuous He gas flow of 1.5 mL min−1 and a
Isolation of taiwachelin
The supernatant from a 12 L culture was separated from the cells
by centrifugation at 9500g. Siderophores that had been secreted
into the culture broth during cultivation were recovered by
adsorption onto 150 g L−1 Amberlite XAD-2 (Supelco) over-
night. After two washing steps with distilled water, the adsorbed
metabolites were eluted from the resin with 3 × 200 mL of
methanol. The eluate was concentrated under vacuum to dryness
prior to resuspension in 2 mL of methanol. Crude taiwachelin
was isolated from this extract on a Shimadzu UFLC liquid
chromatography system equipped with a Nucleodur C18 HTec
column (VP 250 × 10 mm, 5 μm; Macherey-Nagel) using an iso-
cratic flow of 75% methanol in water + 0.1% trifluoroacetic acid
at a flow rate of 3.5 mL min−1 with wavelength monitoring at
210 nm. Final purification was accomplished with successive
HPLC separation on a Nucleodur PFP column (VP 250 ×
10 mm, 5 μm; Macherey-Nagel) using the same conditions as
described above.
Taiwachelin (1). White powder: FT-IR νmax/cm−1 3291, 2924,
1
2845, 1644, 1537. H-NMR (500 MHz, dimethylsulfoxide-d6)
δH [ppm] (J [Hz]) 0.85 (3 H, t, J 7.0, H-41), 1.03 (3 H, d, J 6.4,
H-25), 1.07 (3 H, d, J 6.3, H-17), 1.23 (14 H, m, H-33–H-39),
1.26 (2 H, m, H-40), 1.46 (2 H, m, H-32), 1.47 (2 H, m, H-11),
1.55 (2 H, m, H-12), 1.70 (1 H, m, Hb-4), 1.90 (2 H, m, H-3),
1.92 (1 H, m, Ha-4), 2.14 (2 H, t, J 7.3, H-31), 2.37 (1 H, dd, J
15.0, 6.3, Hb-15), 2.52 (1 H, dd, J 15.0, 6.3, Ha-15), 2.53 (1 H,
dd, J 16.5, 7.2, Hb-20), 2.70 (1 H, dd, J 16.5, 6.8, Ha-20), 3.42
(1 H, m, Hb-13), 3.48 (2 H, m, H-2), 3.50 (1 H, m, Ha-13), 3.59
(2 H, d, J 7.2, H-8), 3.82 (1 H, quint, J 6.4, H-24), 4.00 (1 H,
sext J 6.3, H-16), 4.19 (1 H, dd, J 8.0, 6.4, H-23), 4.29 (1 H, m,
H-10), 4.31 (1 H, m, H-7), 4.34 (1 H, ddd, J 10.7, 8.4, 5.1, H-5),
4.48 (1 H, d J 2.7, H-28), 4.56 (1 H, ddd, J 7.8, 7.2, 6.8, H-19),
4.74 (1 H, dd, J 9.2, 2.7, H-27), 7.73 (1 H, d, J 8.0, NH), 7.85
(1 H, d, J 9.2, NH), 7.91 (1 H, d, J 7.6, NH), 7.95 (1 H, d, J 7.8,
NH), 8.09 (1 H, d, J 8.4, NH), 8.15 (1 H, d, J 7.8, NH).
13C-NMR (125 MHz, dimethylsulfoxide-d6), δC [ppm] 14.0
(C-41), 19.7 (C-25), 20.3 (C-3), 22.2 (C-40), 22.8 (C-12), 23.4
(C-17), 25.3 (C-32), 27.3 (C-4), 28.7 (C-33), 28.8 (C-34), 28.9
(C-38), 29.0 (C-35), 29.1 (C-11), 29.1 (C-36), 29.1 (C-37), 31.4
(C-39), 35.3 (C-31), 36.1 (C-20), 41.5 (C-15), 46.7 (C-13), 49.7
(C-19), 49.9 (C-5), 51.2 (C-2), 52.7 (C-10), 55.1 (C-7), 55.5
(C-27), 58.8 (C-23), 62.1 (C-8), 63.2 (C-16), 67.2 (C-24), 70.5
temperature gradient from 40 to 200 °C at a rate of 5 °C min−1
Co-chromatography of the derivatised taiwachelin
hydrolysate was conducted against a derivatised Hbu standard
in configuration as well as a mixture containing both
.
S
enantiomers.
Photoreactivity of taiwachelin
Reduction of the complexed ferric iron to ferrous iron was
examined using a previously reported protocol.24 Briefly,
each reaction contained 100 μM taiwachelin, 10 μM FeCl3
and 40 μM of the trapping agent BPDS (Fluka) in a PBS
buffer (pH 7.5). The reactions were either exposed to sunlight or
2+
kept in the dark for 4 h. The formation of Fe(BPDS)3 was
recorded before and after exposure to sunlight/darkness by
measuring the absorption at 535 nm using a Genesys 10 UV
spectrophotometer (Thermo). All experiments were run in
triplicate.
9342 | Org. Biomol. Chem., 2012, 10, 9338–9343
This journal is © The Royal Society of Chemistry 2012

View Article Online
10 (a) D. W. Udwary, L. Zeigler, R. N. Asolkar, V. Singan, A. Lapidus,
W. Fenical, P. R. Jensen and B. S. Moore, Proc. Natl. Acad. Sci. U. S. A.,
2007, 104, 10376; (b) K. Penn, C. Jenkins, M. Nett, D. W. Udwary,
E. A. Gontang, R. P. McGlinchey, B. Foster, A. Lapidus, S. Podell,
E. E. Allen, B. S. Moore and P. R. Jensen, ISME J., 2009, 3, 1193.
11 M. S. Thomas, Biometals, 2007, 20, 431.
12 (a) M. F. Kreutzer, H. Kage, P. Gebhardt, B. Wackler, H. P. Saluz,
D. Hoffmeister and M. Nett, Appl. Environ. Microbiol., 2011, 77, 6117;
(b) S. W. Bearden, J. D. Fetherstone and R. D. Perry, Infect. Immun.,
1997, 65, 1659; (c) G. Bhatt and T. P. Denny, J. Bacteriol., 2004, 186,
7896.
13 C. Amadou, G. Pascal, S. Mangenot, M. Glew, C. Bontemps, D. Capela,
S. Carrère, S. Cruveiller, C. Dossat, A. Lajus, M. Marchetti, V. Poinsot,
Z. Rouy, B. Servin, M. Saad, C. Schenowitz, V. Barbe, J. Batut,
C. Médigue and C. Masson-Boivin, Genome Res., 2008, 18, 1472.
14 (a) E. A. Felnagle, J. J. Barkei, H. Park, A. M. Podevels,
M. D. McMahon, D. W. Drott and M. G. Thomas, Biochemistry, 2010,
49, 8815; (b) W. Zhang, J. R. Heemstra Jr., C. T. Walsh and H. J. Imker,
Biochemistry, 2010, 49, 9946; (c) A. H. Asghar, S. Shastri, E. Dave,
I. Wowk, K. Agnoli, A. M. Cook and M. S. Thomas, Microbiology, 2011,
157, 349.
Acknowledgements
We gratefully acknowledge financial support from the Bundesmi-
nisterium für Bildung und Forschung (BMBF GenoMik-Transfer
programme; grant# 0315591A). We thank A. Perner and
T. Neuwirth (HKI Jena, Department of Biomolecular Chemistry)
for recording high resolution ESI-MS data and for assistance
with GC-MS measurements, respectively. We also thank
M. Poetsch (HKI Jena, Department of Molecular and Applied
Microbiology) for MALDI-TOF/TOF measurements.
References
1 (a) J. M. Winter, S. Behnken and C. Hertweck, Curr. Opin. Chem. Biol.,
2011, 15, 22; (b) C. Corre and G. L. Challis, Nat. Prod. Rep., 2009, 26,
977; (c) H. Gross, Appl. Microbiol. Biotechnol., 2007, 75, 267.
2 R. C. Hider and X. Kong, Nat. Prod. Rep., 2010, 27, 637.
3 (a) J. H. Crosa and C. T. Walsh, Microbiol. Mol. Biol. Rev., 2002, 66,
223; (b) G. L. Challis, ChemBioChem, 2005, 6, 601.
4 (a) G. L. Challis, J. Ravel and C. A. Townsend, Chem. Biol., 2000, 7,
211; (b) T. Stachelhaus, H. D. Mootz and M. A. Marahiel, Chem. Biol.,
1999, 6, 493.
15 D. B. Hansen, S. B. Bumpus, Z. D. Aron, N. L. Kelleher and
C. T. Walsh, J. Am. Chem. Soc., 2007, 129, 6366.
16 S. L. Clugston, S. A. Sieber, M. A. Marahiel and C. T. Walsh, Biochemis-
try, 2003, 42, 12095.
17 H. Laatsch, AntiBase: The Natural Compound Identifier, Wiley-VCH,
Weinheim, Germany, 2008.
18 (a) K. Scherlach and C. Hertweck, Org. Biomol. Chem., 2009, 7, 1753;
(b) H. B. Bode, B. Bethe, R. Höfs and A. Zeeck, ChemBioChem, 2002,
3, 619.
5 (a) S. Lautru, R. J. Deeth, L. M. Bailey and G. L. Challis, Nat. Chem.
Biol., 2005, 1, 265; (b) E. J. Dimise, P. F. Widboom and S. D. Bruner,
Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 15311; (c) O. Lazos, M. Tosin,
A. L. Slusarczyk, S. Boakes, J. Cortés, P. J. Sidebottom and P. F. Leadlay,
Chem. Biol., 2010, 17, 160; (d) L. Robbel, T. A. Knappe, U. Linne,
X. Xie and M. A. Marahiel, FEBS J., 2010, 277, 663.
6 M. F. Kreutzer, H. Kage and M. Nett, J. Am. Chem. Soc., 2012, 134,
5415.
7 (a) A. Butler and R. M. Theisen, Coord. Chem. Rev., 2010, 254, 288;
(b) M. Sandy and A. Butler, Chem. Rev., 2009, 109, 4580.
8 (a) A. Powell, M. Borg, B. Amir-Heidari, J. M. Neary, J. Thirlway,
B. Wilkinson, C. P. Smith and J. Micklefield, J. Am. Chem. Soc., 2007,
129, 15182; (b) O. A. Trivedi, P. Arora, V. Sridharan, R. Tickoo,
D. Mohanty and R. S. Gokhale, Nature, 2004, 428, 441;
(c) M. Wittmann, U. Linne, V. Pohlmann and M. A. Marahiel, FEBS J.,
2008, 275, 5343.
9 G. M. Garrity, J. A. Bell and T. Lilburn, in Bergey’s manual of systematic
bacteriology, volume two: the proteobacteria, part C, The Alpha-, Beta-,
Delta-, and Epsilonproteobacteria, ed. G. M. Garrity, D. J. Brenner,
N. R. Krieg and J. T. Staley, Springer, New York, 2nd edn, 2005.
19 V. V. Homann, M. Sandy, J. A. Tincu, A. S. Templeton, B. M. Tebo and
A. Butler, J. Nat. Prod., 2009, 72, 884.
20 B. Schwyn and J. B. Neilands, Anal. Biochem., 1987, 160, 47.
21 (a) E. M. Schaffner, R. Hartmann, K. Taraz and H. Budzikiewicz,
Z. Naturforsch., 1996, 51c, 139; (b) M. J. Dilworth, K. C. Carson,
R. G. F. Giles, L. T. Byrne and A. R. Glenn, Microbiology, 1998, 144,
781; (c) E. Pretsch, P. Bühlmann and C. Affolter, in Structure Determi-
nation of Organic Compounds: Tables of Spectral Data, Springer, Berlin,
3rd edn, 2000.
22 P. Marfey, Carlsberg Res. Commun., 1984, 49, 591.
23 D. R. Goodlett, P. A. Abuaf, P. A. Savage, K. A. Kowalski,
T. K. Mukherje, J. W. Tolan, N. Corkum, G. Goldstein and
J. B. Crowther, J. Chromatogr., A, 1995, 707, 233.
24 S. A. Amin, D. H. Green, M. C. Hart, F. C. Küpper, W. G. Sunda and
C. J. Carrano, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 17071.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 9338–9343 | 9343