A Membrane-Bound Prenyltransferase Catalyzes the O-Prenylation of 1,6-Dihydroxyphenazine in the Marine Bacterium Streptomyces sp. CNQ-509

Article abstract of DOI:10.1002/cbic.201402394

Streptomyces sp. CNQ-509 produces the rare O-prenylated phenazines marinophenazines A and B. To identify the enzyme catalyzing the O-prenyl transfer in marinophenazine biosynthesis, we sequenced the genome of S. sp. CNQ-509. This led to the identification of two genomic loci harboring putative phenazine biosynthesis genes. The first locus contains orthologues for all seven genes involved in phenazine-1-carboxylic acid biosynthesis in pseudomonads. The second locus contains two known phenazine biosynthesis genes and a putative prenyltransferase gene termed cnqPT1. cnqPT1 codes for a membrane protein with sequence similarity to the prenyltransferase UbiA of ubiquinone biosynthesis. The enzyme CnqPT1 was identified as a 1,6-dihydroxyphenazine geranyltransferase, which catalyzes the C-O bond formation between C-1 of the geranyl moiety and O-6 of the phenazine scaffold. CnqPT1 is the first example of a prenyltransferase catalyzing O-prenyl transfer to a phenazine.

Full text of DOI:10.1002/cbic.201402394

CHEMBIOCHEM
FULL PAPERS
DOI: 10.1002/cbic.201402394
A Membrane-Bound Prenyltransferase Catalyzes the
O-Prenylation of 1,6-Dihydroxyphenazine in the Marine
Bacterium Streptomyces sp. CNQ-509
Philipp Zeyhle,[a] Judith S. Bauer,[a] Marco Steimle,[a] Franziska Leipoldt,[a] Manuela Rçsch,[a]
Jçrn Kalinowski,[b] Harald Gross,*[a] and Lutz Heide*[a]
Streptomyces sp. CNQ-509 produces the rare O-prenylated phe-
nazines marinophenazines A and B. To identify the enzyme cat-
alyzing the O-prenyl transfer in marinophenazine biosynthesis,
we sequenced the genome of S. sp. CNQ-509. This led to the
identification of two genomic loci harboring putative phena-
zine biosynthesis genes. The first locus contains orthologues
for all seven genes involved in phenazine-1-carboxylic acid bio-
synthesis in pseudomonads. The second locus contains two
known phenazine biosynthesis genes and a putative prenyl-
transferase gene termed cnqPT1. cnqPT1 codes for a membrane
protein with sequence similarity to the prenyltransferase UbiA
of ubiquinone biosynthesis. The enzyme CnqPT1 was identified
as a 1,6-dihydroxyphenazine geranyltransferase, which catalyz-
es the CÀO bond formation between C-1 of the geranyl
moiety and O-6 of the phenazine scaffold. CnqPT1 is the first
example of a prenyltransferase catalyzing O-prenyl transfer to
a phenazine.
Introduction
Sediment-derived marine actinomycetes are widely recognized
as an emerging source of novel and bioactive secondary me-
tabolites.[1] They have been phylogenetically classified in 13
groups (MAR1–MAR13) within six families.[1a] The MAR4 group
(family Streptomycetaceae) comprises 57 cultured strains[2] and
is known to be a prolific source for a broad range of structural-
ly unprecedented and pharmacologically active hybrid poly-
ketide-terpenoid compounds, such as marinone, azamerone,
and napyradiomycins.[3] One member of the MAR4 group is
Streptomyces sp. CNQ-509, isolated from a marine sediment
sample collected off La Jolla, California, USA. Analysis of this
strain recently led to the isolation of the cytotoxic prenylated
nitropyrroles nitropyrrolins A–E,[4] and O-prenylated phenazines
(Scheme 1), termed marinophenazine A and marinophenazine
B[5] (also known as phenaziterpene A).[6] Such O-prenylated
phenazine ethers are rare and are produced exclusively by
actinomycetes (e.g., phenaziterpenes A and B)[6] and by the ar-
chaeon Methanosarcina mazei Gç1 (methanophenazine).[7]
Previous studies have identified several prenyltransferases
that catalyze the C-prenylation of phenazines in the secondary
Scheme 1. Structures of the prenylated phenazine ethers marinophenazi-
ne A (1) and B (2) from Streptomyces sp. CNQ-509.
metabolism of Streptomyces strains.[8] However, no enzyme has
been identified to catalyze the O-prenylation of phenazines. In
order to elucidate the genetic and biochemical basis of this
reaction, we sequenced the genome of Streptomyces sp. CNQ-
509. This led to the discovery of a complete phenazine biosyn-
thetic gene cluster as well as of an additional genetic locus
containing two putative phenazine biosynthesis genes and a
gene termed cnqPT1. The latter shows similarity to membrane-
bound prenyltransferases of ubiquinone biosynthesis and was
heterologously expressed in Escherichia coli. Biochemical inves-
tigation proved that the gene product converted 1,6-dihy-
droxyphenazine (1,6-DHP) and geranyl diphosphate (GPP) into
marinophenazine B. CnqPT1 also accepted 1-hydroxyphenazine
and the polyketide flaviolin as substrates, thus leading to the
enzymatic formation of two new O-prenylated compounds,
termed marinophenazine C and 2-O-geranylflaviolin, respec-
tively.
[a] P. Zeyhle, J. S. Bauer, M. Steimle, F. Leipoldt, M. Rçsch, Prof. Dr. H. Gross,
Prof. Dr. L. Heide
Pharmazeutische Biologie, Pharmazeutisches Institut
Eberhard Karls Universitꢀt Tꢁbingen
Auf der Morgenstelle 8, 72076 Tꢁbingen (Germany)
E-mail: harald.gross@uni-tuebingen.de
heide@uni-tuebingen.de
[b] Prof. Dr. J. Kalinowski
Microbial Genomics and Biotechnology
Center for Biotechnology, Universitꢀt Bielefeld
Universitꢀtsstrasse 27, 33615 Bielefeld (Germany)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201402394.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2385 – 2392 2385

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
similar to PhzS from Pseudomonas aeruginosa PAO1,[11] and
cnq725 codes for a putative methyltransferase showing similar-
ity to PhzM from the same Pseudomonas strain.[10b] Further-
more, Cnq738 showed similarity to prenyltransferases of the
ABBA superfamily,[12] for example, PpzP and EpzP, the two pre-
nyltransferases of endophenazine biosynthesis.[8a,b] Therefore,
we first speculated that Cnq738 catalyzes O-prenylation in
marinophenazine biosynthesis. In a postulated pathway for the
biosynthesis of marinophenazine A and B (1 and 2; Scheme 2),
2 is a geranylated derivative of 1,6-DHP, and 1,6-DHP is likely
derived from phenazine-1,6-dicarboxylic acid (PDC). PDC is
a key intermediate in the biosynthesis of phenazines in some
Streptomyces strains.[13] In analogy to the reaction catalyzed by
PhzS,[11] PDC might be converted into 1,6-DHP in two hydroxy-
lative decarboxylation steps. Subsequent transfer of a geranyl
moiety (C10) and oxidation of the nitrogen would lead to 1.
To test whether cnq738 indeed codes for the prenyltransfer-
ase in the biosynthesis of 1 and 2, we cloned cnq738 into an
expression vector. The corresponding His-tagged Cnq738 was
overexpressed in E. coli and purified to apparent homogeneity.
Unexpectedly, however, no formation of prenylated products
was observed in incubations of Cnq738 with 1,6-DHP and GPP
(data not shown). Therefore, we searched the genome of Strep-
tomyces sp. CNQ-509 for further putative prenyltransferase and
associated phenazine genes.
Results
Analysis of the Streptomyces sp. CNQ-509 genome sequence
In order to identify the gene cluster for marinophenazine bio-
synthesis, we sequenced the genomic DNA of Streptomyces sp.
CNQ-509 and obtained a draft genome sequence. Analysis on
the antiSMASH 2.0 platform[9] identified a putative phenazine
gene cluster with 26 coding sequences (Figure 1A, Table S1 in
Figure 1. A) Putative phenazine biosynthetic gene cluster and B) the region
around cnqPT1 in the genome sequence of Streptomyces sp. CNQ-509. See
Tables S1 and S2 for further information concerning these loci.
Approximately 169 kb from the phenazine gene cluster, we
identified a locus that consisted of genes for a putative prenyl-
transferase, a putative core phenazine biosynthesis enzyme,
and two putative monooxygenases (Figure 1B, Table S2). Gene
cnq908 showed similarity to phzF from Pseudomonas fluores-
cens which codes for a trans-2,3-dihydro-3-hydroxyanthranilate
isomerase, an essential enzyme in the biosynthesis of phena-
zines.[14] Genes cnq909 and cnq911 encode putative monooxy-
genases; Cnq909 showed similarity to a cyclohexanone mono-
oxygenase, which catalyzes the oxidation of a cyclic ketone
into a lactone.[15] The putative flavin-dependent monooxygen-
ase Cnq911 is similar to PhzS from Pseudomonas aeruginosa.[11]
the Supporting Information). Seven genes (designated cnq726–
731 and cnq734) showed high similarity to the genes of the
phzABCDEFG operon identified in several Pseudomonas
strains.[10] In pseudomonads these genes code for enzymes
that catalyze the formation of phenazine-1-carboxylic acid
(PCA) from intermediates of the shikimate pathway. Near these
core phenazine biosynthesis genes we also identified three
genes coding for putative phenazine-modifying enzymes.
cnq735 codes for a putative flavin-dependent monooxygenase
Scheme 2. Putative biosynthetic pathway of marinophenazine A (1) and B (2) in Streptomyces sp. CNQ-509.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2385 – 2392 2386

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
PhzS catalyzes the oxidative de-
carboxylation of 5-methyl-PCA,
the final step in the biosynthesis
of pyocyanin from PCA. Between
these two putative monooxyge-
nase genes was a gene (desig-
nated cnqPT1) that codes for an
enzyme with similarity to mem-
brane-bound 4-hydroxybenzoate
polyprenyltransferases of ubiqui-
none biosynthesis.[16] CnqPT1
Table 1. Membrane localization of CnqPT1 activity. The reaction mixtures (100 mL) contained Tris·HCl (50 mm,
pH 8.8), 0.5 mm 1,6-DHP, 0.5 mm GPP, 5 mm MgCl2, 20 mm NaCl, and 11.4 mL of each fraction. The mixtures
were incubated for 10 min at 308C.
Fraction
Total protein
[mg]
Product formation
Specific activity
[pmolsÀ1 mgÀ1
[nmolsÀ1
]
[%]
]
[%]
crude protein extract
supernatant (100000g)
membrane fraction
607.8
516.6
75.4
8.9
0.1
5.2
100.0
1.1
58.4
14.6
0.1
69.2
100.0
0.7
474.0
comprises 334 amino acids with six transmembrane helices, as
predicted by the TMHMM server (v. 2.0),[17] thus CnqPT1 was
predicted to be an integral membrane protein.
To verify the subcellular localization of the prenyltransferase
activity we compared the crude protein extract, the super-
natant after centrifugation at 100000g, and the pelleted mem-
brane fraction (Table 1). The enzymatic activity was almost
completely localized in the membrane fraction; only negligible
activity was found in the supernatant. The specific activity of
the membrane fraction was 4.7 times higher than that of the
crude protein extract. These data proved that CnqPT1 is a
membrane-bound enzyme (located in the membrane of E. coli
after heterologous expression). All further biochemical experi-
ments were carried out with the membrane fraction.
Enzyme activity and subcellular localization of CnqPT1
To investigate the biochemical activity of the putative prenyl-
transferase CnqPT1, cnqPT1 was amplified from genomic DNA
of Streptomyces sp. CNQ-509 and ligated into the vector pET-
28a(+). The resulting plasmid pPH25 was expressed in E. coli,
and a crude protein extract was prepared. In order to isolate
the membrane fraction, the crude extract was subjected to
ultracentrifugation at 100000g. The resulting pellet was resus-
pended and incubated with 1,6-DHP, GPP, and Mg2+. HPLC
analysis readily showed the formation of an enzymatic product
(Figure 2A). In a corresponding LC-MS analysis run, a molecular
ion at m/z 349 [M+H]+ was detected for the product, thus
suggesting it to be a monogeranylated derivative of 1,6-DHP
(incubations with a membrane fraction from cells harboring
the empty vector showed no product formation; Figure 2A). A
comparative analysis of the spectral properties of the product
with its reactant (1,6-DHP) as well as an independent structure
elucidation employing 1D and 2D NMR spectroscopy and
high-resolution MS confirmed that the enzymatic product was
marinophenazine B (2; Scheme 1 and Figures S1–S6, Tables S3
and S4).
Biochemical properties of CnqPT1
In our standard assay for prenyltransferase activity (see the
Experimental Section), product formation was linear up to
0.5 mgmLÀ1 membrane protein and up to 20 min incubation
time. Maximum product formation was at pH 8.8. Addition of
20 mm NaCl slightly increased the activity (by ~6%), and there-
fore 20 mm NaCl was added in all subsequent assays.
Dependence of the prenyltransferase activity on divalent
cations
The enzymatic activity of membrane-bound prenyltransferases
of lipoquinone biosynthesis (e.g., ubiquinone biosynthesis) is
Figure 2. HPLC analysis of prenyltransferase assays. Upper chromatograms: membrane fraction from E. coli harboring the empty vector pET-28a(+). Lower
chromatograms: membrane fraction from E. coli harboring the cnqPT1 expression plasmid pPH25. Incubations contained GPP, Mg2+, and aromatic substrates:
A) 1,6-dihydroxyphenazine (1,6-DHP), B) 1-hydroxyphenazine, and C) flaviolin.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2385 – 2392 2387

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
absolutely dependent on the presence of Mg2+ or similar diva-
lent cations.[18] These enzymes contain an aspartate-rich motif
(e.g., NxxxDxxxD) for binding to the prenyl diphosphate with
an Mg2+ ion.[18–19] Likewise, the activity of CnqPT1 was also
strictly dependent on the presence of Mg2+ ions, with the
highest activity detected at 5 mm Mg2+. The addition of 2 mm
EDTA without divalent cations abolished product formation,
thus proving the absolute requirement for divalent cations for
CnqPT1 prenyltransferase activity. Replacement of Mg2+ by
Ca2+ (5 mm) reduced activity by 99%. These results are consis-
tent with the observation that CnqPT1 incorporates the aspar-
tate-rich motif (aa 90–98: NALADREQD).
substrate (maximum velocity 9.2 nmolminÀ1 mgÀ1). Therefore,
the catalytic activity of CnqPT1 for 1,6-DHP was nearly twice as
high as for 1-hydroxyphenazine.
Identification of enzymatically generated marinophenazine
C (3)
The structure of compound 3 was elucidated by comparison of
the spectral properties of the product with that of its aromatic
reactant. Product 3 was formed by the enzymatic reaction of
1-hydroxyphenazine with GPP. In comparison to 1-hydroxyphe-
nazine, the 13C NMR spectrum of 3 showed ten additional reso-
nances (Table 2, Figure S8), attributed to a geranyl moiety by
1H,1H COSY, 1H,13C HSQC, and 1H,13C HMBC NMR experiments.
Further interpretation of the NMR data revealed the regiospeci-
ficity of the prenylation reaction. In comparison with the re-
actant, both spin systems of the phenazine core structure re-
mained mainly intact (Figure S7), thus suggesting that 1-hy-
droxyphenazine was not C-prenylated but prenylated rather at
one of the heteroatoms. 2D NMR experimental results, particu-
larly the 1H,13C HMBC cross correlation between CH2-1’ and
quaternary carbon C-1, confirmed that product 3 was O-preny-
lated (Figure S9). Product 3 was thus determined to be 1-((3,7-
dimethylocta-2E,6-dien-1-yl)oxy)phenazine and named marino-
phenazine C (Scheme 3).
Substrate specificity
To test the substrate specificity of CnqPT1, the enzyme was
incubated with different aromatic and isoprenoid substrates.
With 1,6-DHP as the aromatic substrate, GPP was clearly the
preferred isoprenoid substrate. Use of dimethylallyl diphos-
phate or farnesyl diphosphate resulted in only 0.5 or 1.7% ac-
tivity, respectively, relative to the value observed with GPP.
When CnqPT1 was incubated with 1-hydroxyphenazine and
GPP, formation of 3 was detected by HPLC (Figure 2B). LC-MS
analysis showed a molecular ion at m/z 333 [M+H]+ for com-
pound 3, thus suggesting it to be a monoprenylated derivative
of 1-hydroxyphenazine (the exact structure of 3 is described
below). The reaction with 1-hydroxyphenazine and dimethylall-
yl diphosphate revealed 1.0% activity relative to that with
GPP; with 1-hydroxyphenazine and farnesyl diphosphate it was
1.8%. No product formation was observed when CnqPT1 was
incubated with the flavonoid naringenin or with 1,6-dihydroxy-
naphthalene. However, incubations with flaviolin (2,5,7-trihy-
droxy-1,4-naphthoquinone) and GPP resulted in the formation
of 4 and 5 (Figure 2C). In LC-MS analyses both products
showed a molecular ion at m/z 341 [MÀH]À, thus suggesting
that they are monogeranylated derivatives of flaviolin (struc-
ture elucidation of the main product 4 is explained below). In
comparison to the presumed genuine substrate 1,6-DHP, the
enzymatic geranylation of flaviolin proceeded at least six times
slower. 4-Hydroxybenzoate is the substrate of the prenylation
reaction in ubiquinone biosynthesis in E. coli.[16] Incu-
Identification of enzymatically generated 2-O-geranyl-
flaviolin (4)
As described above, incubation of CnqPT1 with flaviolin and
GPP yielded the two prenylated products 4 and 5. Although
we were able to obtain a sufficient amount of 4 for structure
elucidation by a preparative-scale assay (see the Experimental
Section), the low amount of 5 prevented complete elucidation
of its structure.
A complete set of NMR spectra revealed that 4 is an O-pre-
nylated flaviolin derivative (Table 3, Scheme 3). The 1H NMR
spectrum exhibited three resonances, which were attributed to
the methine groups of the flaviolin core, thus indicating O-pre-
nylation of flaviolin (Figure S10). Interpretation of the 2D NMR
bations with 4-hydroxybenzoate and GPP or farnesyl
diphosphate showed low product formation, but
these products were also detected in incubations
with membrane fractions harboring the empty
vector, and therefore are most likely due to the UbiA
homologue of the E. coli expression strain.
Steady-state kinetics of CnqPT1
The CnqPT1 reaction displayed Michaelis–Menten ki-
netics. For 1,6-DHP and GPP (the presumed genuine
substrates) the apparent Km values were 255 and
47 mm, respectively; the maximum velocity of this re-
action was 34.0 nmolminÀ1 mgÀ1. With 1-hydroxyphe-
nazine as the aromatic substrate, an apparent Km of
131 mm was determined with GPP as the isoprenoid Scheme 3. Enzymatic products marinophenazine C (3) and 2-O-geranylflaviolin (4).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2385 – 2392 2388

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
derivatives,
strikingly,
both
Table 2. NMR spectroscopic data for 1-hydroxyphenazine and marinophenazine C (3) in [D6]acetone (d in ppm,
J in Hz).
locus A and locus B contain
a gene encoding a prenyltransfer-
ase of a different class (cnq738
Position
1-Hydroxyphenazine
Marinophenazine C (3)
[a,b]
[b,c,d]
[e]
[d,f]
dH
dC
dH
dC
HMBC[g]
and
cnqPT1,
respectively).
1
2
3
153.7 qC
110.1 CH
132.7 CH
120.5 CH
144.8 qC
144.9[i] qC
130.5[j] CH
131.4[h] CH
131.7[h] CH
130.1[j] CH
142.1[i] qC
136.0 qC
155.6 qC
cnq738 codes for a putative pre-
nyltransferase of the ABBA su-
perfamily,[12,20] and was initially
suspected to be responsible for
the prenylation reaction because
of the resemblance to known
gene cluster structures.[8] To our
surprise, Cnq738 was unable to
catalyze the expected prenyla-
tion reaction in marinophena-
zine biosynthesis. The physiolog-
ical function, if any, of Cnq738
remains unknown. Intriguingly,
cnqPT1 is completely unrelated
to ABBA prenyltransferases, and
is predicted to be a membrane-
bound aromatic prenyltransfer-
ase. Biochemical investigations
proved that CnqPT1 transfers
a geranyl moiety onto 1,6-DHP
to form 2. As 2 was also isolated
from Streptomyces sp. CNQ-
509,[5] it represents a likely inter-
7.23 (dd, J=7.4, 1.2)
7.83 (dd, J=8.9, 7.4)
7.73 (dd, J=8.9, 1.2)
7.23 (dd, J=7.4, 1.2)
7.80 (dd, J=8.8, 7.4)
7.75 (dd, J=8.8, 1.2)
109.1 CH
131.7 CH
121.8 CH
145.2 qC
143.0[i] qC
130.2[j] CH
131.6[h] CH
130.9[h] CH
130.8[j] CH
144.1[i] qC
138.2 qC
66.8 CH2
120.6 CH
141.5 qC
40.2 CH2
27.0 CH2
124.7 CH
132.1 qC
17.7 CH3
25.8 CH3
16.8[k] CH3
1, 4, 10a
1, 2, 4a, 10a
2, 10a
4
4a
5a
6
7
8
8.22[j] (m)
7.93[h] (m)
7.93[h] (m)
8.22[j] (m)
8.19[j] (d, J=7.9)
7.91[h] (m)
5a, 8
5a, 6, 9a
5a, 6, 9a
7, 9a
7.91[h] (m)
9
8.26[j] (d, J=7.9)
9a
10a
1’
2’
3’
4’
5’
6’
7’
8’
9’
10’
OH
4.91 (d, J=6.4)
5.67 (t, J=6.4)
1, 2’, 3’
4’, 10’
2.14 (m)
2.14 (m)
5.12 (m)
2’, 3’, 5’
4’, 6’, 7’
5’
1.59 (s)
1.63 (s)
1.82 (s)
6’, 7’, 9’
6’, 7’, 8’
2’, 3’, 4’
9.23 (br)
[a] Recorded at 400 MHz. [b] Measured values were in good agreement with published NMR data for 1-hy-
droxy- or 1-methoxyphenazine.[34,36] [c] Recorded at 101 MHz. [d] Multiplicity determined by a multiplicity
edited 1H,13C HSQC NMR experiment. [e] Recorded at 600 MHz. [f] Recorded at 151 MHz. [g] Protons showing
long-range correlation with indicated carbon. [h], [i], [j] Assignments interchangeable. [k] On the basis of the
13C NMR chemical shift for C-10’ (dC =16.8), the geometry of the D2’,3’ double bond was defined as E.[37]
Table 3. NMR spectroscopic data for 2-O-geranylflaviolin (4) in
[D6]acetone (d in ppm, J in Hz).
data (Figures S11–S14) further supported this assumption and
revealed the regiospecificity of the prenylation reaction. Cross
1
correlations in the H,1H ROESY NMR spectrum between CH2-1’
[a]
[b,c]
Position dH
dC
COSY ROESY
HMBC[d]
1
and CH-3 (Figure S13), as well as the H,13C HMBC correlation
1
2
3
4
4a
5
179.9 qC
160.9 qC
110.8 CH
190.6 qC
108.8[e] qC
164.3 qC
between CH2-1’ and C-2 (Figure S14), unequivocally proved
that flaviolin was O-prenylated at the hydroxy group at posi-
tion 2 of the flaviolin core. Thus, 4 was determined to be 2-
((3,7-dimethylocta-2E,6-dien-1-yl)oxy)-5,7-dihydroxynaphtha-
lene-1,4-dione (Scheme 3) and named 2-O-geranylflaviolin.
6.15 (s)
1’, 5-OH 1, 2, 4, 4a, 5, 8a
6
7
8
8a
6.62 (d, J=2.3) 108.8 CH
164.6 qC
7.07 (d, J=2.3) 108.6 CH
134.1 qC
8
6
5-OH
4a, 5, 7, 8
1, 4a, 6, 7, 8a
Discussion
1’
2’
3’
4’
5’
6’
7’
8’
9’
4.70 (d, J=6.6) 67.2 CH2
5.50 (t, J=6.6) 118.6 CH
143.6 qC
2’
1’
3, 2’,10’ 2, 2’, 3’
Only a few O-prenylated phenazines have been isolated from
microorganisms.[5–7] The unusual O-prenylation of 1 and 2
drew our attention to the corresponding prenylating enzyme.
Analysis of the genome sequence identified the putative gene
cluster of marinophenazines, which possess three distinctive
features. 1) In contrast to most biosynthetic gene clusters of
known prenylated phenazine derivatives,[8] the cnq genes are
not physically linked but are in two separate clusters (Figure 1),
here referred to as loci A and B. Such a split phenazine gene
cluster has so far only been reported for the endophenazine
gene cluster of Streptomyces cinnamonensis.[8b] 2) The genes for
isoprenoid biosynthesis via the mevalonate pathway are also
absent from both loci, in contrast to previously identified gene
clusters for prenylated phenazine biosynthesis.[8] 3) Compared
to all other biosynthetic gene clusters of prenylated phenazine
1’
1’, 4’, 10’
2.12 (m)
2.14 (m)
40.1 CH2
26.9 CH2
5’
4’, 6’
5’
2’, 3’, 5’, 6’, 10’
3’, 4’, 6’, 7’
8’, 9’
5.11 (t, J=6.3) 124.6 CH
132.2 qC
1.59 (s)
1.64 (s)
1.78 (s)
12.50 (s)
17.7 CH3
25.8 CH3
16.7[f] CH3
6’, 7’, 9’
6’, 7’, 8’
2’, 3’, 4’
5, 6, 7
10’
5-OH
1’
3, 6
[a] Recorded at 500 MHz. [b] Recorded at 125 MHz. [c] Multiplicity deter-
mined by multiplicity edited 1H,13C HSQC and 13C JMOD NMR experi-
ments. [d] Protons showing long-range correlation with indicated carbon.
[e] Extracted from the H,13C HMBC spectrum, because of overlapping res-
onances in the 13C JMOD NMR experiment. [f] On the basis of the
13C NMR chemical shift for C-10’ (dC =16.7), the geometry of the D2’,3’
double bond was defined as E.[37]
1
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2385 – 2392 2389

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
nologies, Madison, WI). UV spectra were obtained on a Lambda25
spectrometer (PerkinElmer). 1D and 2D NMR spectra were recorded
on Avance III 400, 500, and 600 and DRX-250 spectrometers
(Bruker). Spectra were referenced to residual solvent signals
([D6]acetone or [D]chloroform; dH/C 2.04/29.8 and 7.27/77.0, respec-
tively). High-resolution mass spectra were acquired on an HR-ESI-
TOF-MS maXis 4G mass spectrometer (Bruker) and on an APEX II
FT-ICR-MS instrument (Bruker).
mediate in the biosynthesis of 1. The isolation of 2 indicates
that the prenylation reaction occurs before N-oxidation rather
than after (Scheme 2). It appears likely that the phenazine
genes of locus A synthesize the phenazine precursor (probably
PDC), whereas the gene products of locus B are involved in
the conversion of this phenazine precursor to
(Scheme 2).
1
The biochemical results obtained for CnqPT1 strongly sug-
gest its involvement in marinophenazine biosynthesis. Howev-
er, confirmation of this hypothesis by genetic experiments was
not possible. Despite repeated attempts, Streptomyces sp.
CNQ-509 could not be manipulated genetically, thus prevent-
ing inactivation of cnqPT1.
Bacterial strains and culture conditions: Streptomyces sp. CNQ-
509 was kindly provided by William Fenical and Paul R. Jensen
(Scripps Institution of Oceanography, University of California, San
Diego) and cultured as described previously.[4] Brevibacterium iodi-
num DSM 433 was obtained from the Deutsche Sammlung von
Mikroorganismen und Zellkulturen GmbH (DSMZ; Braunschweig,
Germany) and cultivated as recommended by the DSMZ. E. coli
XL1 Blue MRF’ (Stratagene/Agilent Technologies) was used for clon-
ing and grown on solid or in liquid LB medium at 378C. Recombi-
nant strains were selected with kanamycin (50 mgmLÀ1).
In contrast to the soluble and mostly Mg2+-independent
ABBA prenyltransferases, CnqPT1 is membrane-bound and
Mg2+ dependent, similar to membrane-bound aromatic prenyl-
transferases of primary metabolism, for example UbiA of ubiq-
uinone biosynthesis and MenA of menaquinone biosynthesis
in E. coli.[16,21] These membrane-bound enzymes were initially
identified in microorganisms solely in primary metabolism,[18]
but recently some of these enzymes were found to be in-
volved in the biosynthesis of secondary metabolites in fungi[22]
and bacteria.[8c,23] In the rare actinomycete Actinoplanes missou-
riensis, a membrane-bound prenyltransferase catalyzes O-pre-
nylation in the biosynthesis of a polyketide.[24] Similarly to
CnqPT1, this enzyme uses GPP as the prenyl donor, but with
a hydroquinone instead of a phenazine as the prenyl acceptor.
Enzymes involved in the O-prenylation of a phenazine scaffold
have not previously been described (either from the ABBA su-
perfamily or from membrane-bound aromatic prenyltransferas-
es). Therefore, CnqPT1 is the first enzyme to catalyze this type
of prenylation reaction. In contrast to UbiA of ubiquinone bio-
synthesis, the prenyl donor specificity of CnqPT1 can be
judged tentatively as rather narrow (GPP as the isoprenoid
substrate). In this context it is noteworthy that CnqPT1 shows
a new regiospecificity regarding the tri-hydroxylated aromatic
compound flaviolin. Previously described prenylated flaviolins
are solely either O-prenylated at position 7 (e.g., the neuropro-
tective flaviogeranin)[25] or C-prenylated at position 3[26] or 6.[27]
As CnqPT1 performs prenylation at position 2 of the flaviolin
skeleton, this prenyltransferase represents also a useful tool to
generate new bioactive prenylated aromatic compounds.
Genetic procedures: DNA isolation and manipulation were carried
out by standard methods (Sambrook and Russell[30] and Kieser
et al.).[31] Genomic-tip 100/G columns (Qiagen) were used to purify
genomic DNA.
Genome sequencing and bioinformatic sequence analysis: An 8k
paired-end and a whole-genome shotgun library were prepared
according to standard protocols (Roche Applied Science). For se-
quencing of Streptomyces sp. CNQ-509 genomic DNA, a Genome
Sequencer FLX System and Titanium chemistry (Roche Applied Sci-
ence) were used. Assembly of the sequence reads was performed
with GS Assembler Software (version 2.5.3; Roche). BLAST[32] was
used for database searches, and the Artemis[33] program was used
for sequence analysis.
Preparation of 1,6-DHP: B. iodinum was precultured for four days
at 308C and 200 rpm in DSMZ medium 1 (50 mL; peptone (0.5%)
and meat extract (0.3%); pH 7.0 before sterilization). The produc-
tion medium (1.25 L; yeast extract (1%), glucose (1%), l-valine
(0.1%), and Diaion HP-20 (1%; Mitsubishi Chemical); pH 7.0 before
sterilization) was inoculated with preculture (20 mL) and cultured
for eight days at 268C and 200 rpm. Subsequently, the culture
broth was extracted with CH2Cl2 (1.25 L). The extract was passed
through a filter paper, dried over Na2SO4, and evaporated, in order
to obtain an iodinin extract (250 mg). Iodinin (1,6-dihydroxyphena-
zine-5,10-dioxide) was reduced to 1,6-DHP according to the
method of Breitmaier and Hollstein.[34] The crude reaction product
1,6-DHP (220 mg) was purified on a silica gel 60 column (mobile
phase: toluene/MeOH (975:25)) to yield 1,6-DHP (31.6 mg).
Protein expression and preparation of enzyme extracts: Primers
cnqpt1_FW2 (5’-CGATCCCATGGGACGACGAGAG-3’) and cnqpt1_
RV2 (5’-GTGGGAATTCGCCGGGTGGTC-3’) were used to amplify the
putative prenyltransferase gene cnqPT1 from Streptomyces sp.
CNQ-509 genomic DNA (NcoI and EcoRI restriction sites under-
lined). The PCR product was cloned into pET-28a(+) (Novagen/
Merck Millipore) to produce the expression construct pPH25,
which was verified by restriction mapping and sequencing.
Experimental Section
Chemicals: Dimethylallyl diphosphate, GPP, and farnesyl diphos-
phate were synthesized according to Woodside et al.[28] Flaviolin
was prepared as described by Gross et al.[29] Other chemicals and
molecular biological agents were obtained from standard commer-
cial sources.
General experimental procedures: HPLC analysis was performed
on a 1100 series system (Agilent Technologies) coupled to a photo-
diode array detector or a variable wavelength detector. ESI-LC/MS
analysis was carried out with a 1200 series system (Agilent Technol-
ogies) coupled to an ESI spectrometer (LC/MSD Ultra Trap System
XCT 6330). FT-IR spectra were recorded on an FT/IR-4200 spectrom-
eter (Jasco) interfaced to a MIRacle ATR (ZnSe) system (PIKE Tech-
An overnight culture (17.5 mL) of E. coli Rosetta 2(DE3)pLysS (Nova-
gen) harboring pPH25 in LB medium with kanamycin (50 mgmLÀ1
)
and chloramphenicol (25 mgmLÀ1) was used to inoculate Terrific
Broth medium[30] (500 mL) supplemented with kanamycin
(50 mgmLÀ1) and chloramphenicol (25 mgmLÀ1). The cells were cul-
tured at 378C and 250 rpm to OD600 =0.6–0.7. After lowering the
temperature to 208C, isopropyl b-d-1-thiogalactopyranoside (IPTG)
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2385 – 2392 2390

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
was added (final concentration, 0.5 mm). The cells were grown
overnight (15–16 h) at 208C and 250 rpm then harvested by centri-
fugation (6080g, 10 min). The resulting pellet was washed with
Tris·HCl (50 mm, pH 7.5) and centrifuged again. Cells were resus-
pended in lysis buffer (Tris·HCl (50 mm, pH 7.5), lysozyme
(0.5 mgmLÀ1), 1,4-dithiothreitol (10 mm); 2 mL per gram cell wet
weight) and ruptured by sonication (W-250D Sonifier; Branson,
Danbury, CT). The lysate was centrifuged (5000g, 15 min, 48C) to
remove cell debris. The obtained extract was passed through PD-
10 desalting columns (GE Healthcare) and eluted with Tris·HCl
(50 mm, pH 7.5). This desalted extract was used as crude protein
extract in measurements of enzyme activity. For isolation of the
membrane fraction the extract was centrifuged (100000g, 75 min,
48C), and the resulting pellet was resuspended in Tris·HCl (50 mm,
pH 7.5) and centrifuged again (100000g, 75 min, 48C). The pellet
was resuspended in Tris·HCl (50 mm, pH 7.5) and used to deter-
mine activity in the membrane fraction. Enzyme activity in the
supernatant fraction was determined by using the supernatant
after the first centrifugation (100000g). Protein concentration was
quantified by a Bradford assay.[35]
(4.73), 371 nm (3.41); IR (ATR): n˜ =2964, 2924, 2856, 1530, 1485,
1358, 1125, 802, 738 cmÀ1 1H and 13C NMR spectroscopic data
;
measured in [D]chloroform (Table S4) and [D6]acetone (Table S3);
HR-ESI-MS: m/z 371.1733 [M+Na]+ (calcd for [C22H24N2O2Na]+:
371.1730, D=+0.8 ppm).
Marinophenazine C, 1-((3,7-dimethylocta-2E,6-dien-1-yl)oxy)phe-
nazine, 3: Yellow powder; UV(MeOH): lmax (loge)=261 (4.61),
365 nm (3.80); IR (ATR): n˜ =2965, 2922, 2854, 1520, 1479, 1087,
1
761, 741 cmÀ1; H and 13C NMR spectroscopic data in Table 2; HR-
ESI-TOF-MS: m/z 333.1963 [M+H]+ (calcd for [C22H25N2O]+:
333.1961, D=+0.6 ppm).
2-O-Geranylflaviolin,
2-((3,7-dimethylocta-2E,6-dien-1-yl)oxy)-
5,7-dihydroxynaphthalene-1,4-dione, 4: Orange powder; UV-
(MeOH): lmax (loge)=216 (4.17), 263 (3.81), 303 nm (3.60); IR (ATR):
1
n˜ =2926, 2361, 1629, 1237, 794 cmÀ1; H and 13C NMR spectroscop-
ic data in Table 3; HR-ESI-MS: m/z 341.1385 [MÀH]À (calcd for
[C20H21O5]À: 341.1384, D=+0.3 ppm).
Nucleotide sequence accession numbers: The nucleotide sequen-
ces of the two genomic loci reported in this study are available in
the GenBank database under accession numbers KJ451627 and
KJ451628.
Assays for prenyltransferase activity: Standard enzyme assays
(100 mL) were performed in Tris·HCl (50 mm, pH 8.8) containing 1,6-
DHP (0.5 mm), GPP (0.5 mm), MgCl2 (5 mm), NaCl (20 mm), and
membrane protein (0.25 mgmLÀ1) for 10 min at 308C. The reaction
mixture was extracted with ethyl acetate/formic acid (100 mL;
975:25). The organic layer was evaporated to dryness, and the
dried residue was dissolved in MeOH for HPLC and LC-MS analysis.
HPLC analysis was performed with an Eclipse XDB-C18 column
(4.6ꢁ150 mm, 5 mm; flow rate 1 mLminÀ1; Agilent Technologies). A
linear gradient (40–100% solvent B over 12 min) with an additional
5 min of 100% solvent B was used (solvent A: H2O/formic acid
(999:1), solvent B: MeOH/formic acid (999:1)). UV detection was
carried out at 270 nm (1-hydroxyphenazine), 275 nm (1,6-DHP),
306 nm (flaviolin), or 370 nm (1,6-DHP and 1-hydroxyphenazine).
LC-MS analysis was performed with a Nucleosil 100 C18 column
(100ꢁ2 mm, 3 mm; flow rate 0.4 mLminÀ1; Dr. Maisch GmbH, Am-
merbuch-Entringen, Germany) at 408C. A linear gradient (40–100%
solvent B over 12 min) was used with an additional 5 min of 100%
solvent B (solvent A: H2O (0.1% formic acid), solvent B: MeOH
(0.06% formic acid)). UV detection was carried out at 260, 275, 305,
and 370 nm. Electrospray ionization (negative and positive) was
performed in ultrascan mode (capillary voltage 3.5 kV, 3508C).
Acknowledgements
We gratefully acknowledge William Fenical and Paul R. Jensen
(University of California, San Diego, USA) for providing the strain
Streptomyces sp. CNQ-509. The authors thank Andreas Kulik
(University of Tꢁbingen, Germany) for LC-MS measurements,
Peter Keck (University of Tꢁbingen, Germany) for help with the
preparation of 1,6-DHP, and Kerstin Seeger, Tobias Bonitz, Ute
Metzger, and Elisa Haug-Schifferdecker for the synthesis of dime-
thylallyl diphosphate, GPP, and farnesyl diphosphate. This study
was supported by grants from the German Federal Ministry of
Education and Research (GenBioCom) to L.H. and J.K.
Keywords: biosynthesis · enzyme catalysis
prenyltransferase · Streptomyces
· phenazine ·
Enzymatic production and isolation of marinophenazine B (2),
marinophenazine C (3), and 2-O-geranylflaviolin (4): Incubations
(25–175 mL scale) were carried out in Tris·HCl (50 mm, pH 8.8) con-
taining the respective aromatic substrate (1 mm), GPP (1 mm), NaCl
(20 mm), MgCl2 (5 mm), and membrane protein (2 mgmLÀ1 (2) or
1 mgmLÀ1 (3, 4)). The reaction mixtures were incubated for 4 h at
308C and extracted with ethyl acetate/formic acid (975:25). Ethyl
acetate layers were evaporated, and the residue was dissolved in
MeOH. The products were purified on a Multospher 120 RP 18HP
column (8ꢁ250 mm, 5 mm; Ziemer Chromatographie, Langerwehe,
Germany). Isocratic elution was performed with 85% (2, 4) or 90%
(3) solvent B for 20 min (2, 3) or 25 min (4) (solvent A: H2O/formic
acid (999:1), solvent B: MeOH/formic acid (999:1)) at a flow rate of
2.5 mLminÀ1. The purification yielded 2.5 mg (2), 5.7 mg (3), and
4.0 mg (4). Compound 4 was further purified by isocratic elution
with acetonitrile (85% in H2O) for 15 min using a Kinetex PFP
column (4.6ꢁ250 mm, 5 mm; Phenomenex, Aschaffenburg, Germa-
ny) at a flow rate of 1 mLminÀ1 to yield 2.4 mg of 4.
[1] a) W. Fenical, P. R. Jensen, Nat. Chem. Biol. 2006, 2, 666–673; b) C. C.
Hughes, W. Fenical, Chem. Eur. J. 2010, 16, 12512–12525.
[2] K. A. Gallagher, K. Rauscher, L. Pavan Ioca, P. R. Jensen, Appl. Environ. Mi-
crobiol. 2013, 79, 6894–6902.
[3] a) C. Pathirana, P. R. Jensen, W. Fenical, Tetrahedron Lett. 1992, 33,
7663–7666; b) J. Y. Cho, H. C. Kwon, P. G. Williams, P. R. Jensen, W. Feni-
cal, Org. Lett. 2006, 8, 2471–2474; c) I. E. Soria-Mercado, A. Prieto-Davo,
P. R. Jensen, W. Fenical, J. Nat. Prod. 2005, 68, 904–910.
[4] H. C. Kwon, A. P. D. M. Espindola, J.-S. Park, A. Prieto-Davꢂ, M. Rose, P. R.
Jensen, W. Fenical, J. Nat. Prod. 2010, 73, 2047–2052.
[5] A. P. D. d. M. Espindola, Ph.D. Thesis, University of California, San Diego
(USA), 2008.
[6] Y. Song, H. Huang, Y. Chen, J. Ding, Y. Zhang, A. Sun, W. Zhang, J. Ju, J.
Nat. Prod. 2013, 76, 2263–2268.
[7] H.-J. Abken, M. Tietze, J. Brodersen, S. Bꢃumer, U. Beifuss, U. Deppen-
meier, J. Bacteriol. 1998, 180, 2027–2032.
[8] a) O. Saleh, B. Gust, B. Boll, H.-P. Fiedler, L. Heide, J. Biol. Chem. 2009,
284, 14439–14447; b) K. Seeger, K. Flinspach, E. Haug-Schifferdecker, A.
Kulik, B. Gust, H.-P. Fiedler, L. Heide, J. Microb. Biotechnol. 2011, 4, 252–
262; c) P. Zeyhle, J. S. Bauer, J. Kalinowski, K. Shin-ya, H. Gross, L. Heide,
PLoS One 2014, 9, e99122; d) D. Heine, K. Martin, C. Hertweck, J. Nat.
Prod. 2014, 77, 1083–1087.
Marinophenazine B, 6-((3,7-dimethylocta-2E,6-dien-1-yl)oxy)-1-
hydroxyphenazine, 2: Yellow powder; UV(MeOH): lmax (loge)=272
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2385 – 2392 2391

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
[9] K. Blin, M. H. Medema, D. Kazempour, M. A. Fischbach, R. Breitling, E.
Takano, T. Weber, Nucleic Acids Res. 2013, 41, W204–W212.
[10] a) D. V. Mavrodi, V. N. Ksenzenko, R. F. Bonsall, R. J. Cook, A. M. Boronin,
L. S. Thomashow, J. Bacteriol. 1998, 180, 2541–2548; b) D. V. Mavrodi,
R. F. Bonsall, S. M. Delaney, M. J. Soule, G. Phillips, L. S. Thomashow, J.
Bacteriol. 2001, 183, 6454–6465.
[11] B. T. Greenhagen, K. Shi, H. Robinson, S. Gamage, A. K. Bera, J. E. Ladner,
J. F. Parsons, Biochemistry 2008, 47, 5281–5289.
[12] a) M. Tello, T. Kuzuyama, L. Heide, J. P. Noel, S. B. Richard, Cell. Mol. Life
Sci. 2008, 65, 1459–1463; b) T. Bonitz, V. Alva, O. Saleh, A. N. Lupas, L.
Heide, PLoS One 2011, 6, e27336.
[23] a) E. Stec, D. Pistorius, R. Mꢄller, S.-M. Li, ChemBioChem 2011, 12, 1724–
1730; b) T. Kawasaki, T. Kuzuyama, K. Furihata, N. Itoh, H. Seto, T. Dairi, J.
Antibiot. 2003, 56, 957–966; c) W. Kitagawa, T. Ozaki, T. Nishioka, Y.
Yasutake, M. Hata, M. Nishiyama, T. Kuzuyama, T. Tamura, ChemBioChem
2013, 14, 1085–1093.
[24] T. Awakawa, N. Fujita, M. Hayakawa, Y. Ohnishi, S. Horinouchi, ChemBio-
Chem 2011, 12, 439–448.
[25] Y. Hayakawa, Y. Yamazaki, M. Kurita, T. Kawasaki, M. Takagi, K. Shin-ya, J.
Antibiot. 2010, 63, 379–380.
[26] Y. Haagen, I. Unsçld, L. Westrich, B. Gust, S. B. Richard, J. P. Noel, L.
Heide, FEBS Lett. 2007, 581, 2889–2893.
[13] M. McDonald, B. Wilkinson, C. W. Van’t Land, U. Mocek, S. Lee, H. G.
Floss, J. Am. Chem. Soc. 1999, 121, 5619–5624.
[27] T. Kumano, T. Tomita, M. Nishiyama, T. Kuzuyama, J. Biol. Chem. 2010,
285, 39663–39671.
[14] W. Blankenfeldt, A. P. Kuzin, T. Skarina, Y. Korniyenko, L. Tong, P. Bayer, P.
Janning, L. S. Thomashow, D. V. Mavrodi, Proc. Natl. Acad. Sci. USA 2004,
101, 16431–16436.
[15] I. A. Mirza, B. J. Yachnin, S. Wang, S. Grosse, H. Bergeron, A. Imura, H.
Iwaki, Y. Hasegawa, P. C. K. Lau, A. M. Berghuis, J. Am. Chem. Soc. 2009,
131, 8848–8854.
[16] M. Melzer, L. Heide, Biochim. Biophys. Acta Lipids Lipid Metab. 1994,
1212, 93–102.
[17] A. Krogh, B. Larsson, G. von Heijne, E. L. L. Sonnhammer, J. Mol. Biol.
2001, 305, 567–580.
[28] A. B. Woodside, Z. Huang, C. D. Poulter, Org. Synth. 1988, 66, 211.
[29] F. Gross, N. Luniak, O. Perlova, N. Gaitatzis, H. Jenke-Kodama, K. Gerth,
D. Gottschalk, E. Dittmann, R. Mꢄller, Arch. Microbiol. 2006, 185, 28–38.
[30] J. Sambrook, D. W. Russel, Molecular Cloning. A Laboratory Manual,
3edrd edCold Spring Harbor Laboratory Press, New York, 2001.
[31] T. Kieser, M. J. Bibb, M. J. Buttner, K. F. Chater, D. A. Hopwood, Practical
Streptomyces Genetics, John Innes Foundation, Norwich, 2000.
[32] S. F. Altschul, T. L. Madden, A. A. Schꢃffer, J. Zhang, Z. Zhang, W. Miller,
D. J. Lipman, Nucleic Acids Res. 1997, 25, 3389–3402.
[33] K. Rutherford, J. Parkhill, J. Crook, T. Horsnell, P. Rice, M.-A. Rajandream,
B. Barrell, Bioinformatics 2000, 16, 944–945.
[34] E. Breitmaier, U. Hollstein, J. Org. Chem. 1976, 41, 2104–2108.
[35] M. M. Bradford, Anal. Biochem. 1976, 72, 248–254.
[18] L. Heide, Curr. Opin. Chem. Biol. 2009, 13, 171–179.
[19] L. Brꢃuer, W. Brandt, D. Schulze, S. Zakharova, L. Wessjohann, ChemBio-
Chem 2008, 9, 982–992.
[20] G. Zocher, O. Saleh, J. B. Heim, D. A. Herbst, L. Heide, T. Stehle, PLoS One
2012, 7, e48427.
[21] K. Suvarna, D. Stevenson, R. Meganathan, M. E. S. Hudspeth, J. Bacteriol.
1998, 180, 2782–2787.
[22] a) T. Itoh, K. Tokunaga, Y. Matsuda, I. Fujii, I. Abe, Y. Ebizuka, T. Kushiro,
Nat. Chem. 2010, 2, 858–864; b) J. Hu, H. Okawa, K. Yamamoto, K.
Oyama, M. Mitomi, H. Anzai, J. Antibiot. 2011, 64, 221–227; c) T. Itoh, K.
Tokunaga, E. K. Radhakrishnan, I. Fujii, I. Abe, Y. Ebizuka, T. Kushiro,
ChemBioChem 2012, 13, 1132–1135; d) H.-C. Lo, R. Entwistle, C.-J. Guo,
M. Ahuja, E. Szewczyk, J.-H. Hung, Y.-M. Chiang, B. R. Oakley, C. C. C.
Wang, J. Am. Chem. Soc. 2012, 134, 4709–4720.
[36] A. Rçmer, Org. Magn. Reson. 1982, 19, 66–68.
[37] a) P. A. Couperus, A. D. H. Clague, J. P. C. M. van Dongen, Org. Magn.
Reson. 1976, 8, 426–431; b) H.-O. Kalinowski, S. Berger, S. Braun,
Carbon-13 NMR Spectroscopy, Wiley, Chichester, 1988; c) M. Albericci,
J. C. Braekman, D. Daloze, B. Tursch, J. P. Declercq, G. Germain, M. van
Meerssche, Bull. Soc. Chim. Belg. 1978, 87, 487–492.
Received: July 20, 2014
Published online on September 15, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 2385 – 2392 2392