Unveiling Two Consecutive Hydroxylations: Mechanisms of Aromatic Hydroxylations Catalyzed by Flavin-Dependent Monooxygenases for the Biosynthesis of Actinorhodin and Related Antibiotics

Article abstract of DOI:10.1002/cbic.201900490

Flavin-dependent monooxygenases are ubiquitous in living systems and are classified into single- or two-component systems. Actinorhodin, produced by Streptomyces coelicolor, is a representative polycyclic polyketide that is hydroxylated through the action of the two-component ActVA-5/ActVB hydroxylase system. These homologous systems are widely distributed in bacteria, but their reaction mechanisms remain unclear. This in vitro investigation has provided chemical proof of two consecutive hydroxylations via hydroxynaphthalene intermediates involved in actinorhodin biosynthesis. The ActVA-5 oxygenase component catalyzed a stepwise dihydroxylation of the substrate, whereas the ActVB flavin reductase not only supplied a reduced cofactor, but also regulated the quinone–hydroquinone interconversion of an intermediate. Our study provides clues for understanding the general biosynthetic mechanisms of highly functionalized aromatic natural products with structural diversity.

Full text of DOI:10.1002/cbic.201900490

Accepted Article
Title: Unveiling Two Consecutive Hydroxylations: Mechanism of
Aromatic Hydroxylations Catalyzed by Flavin-Dependent
Monooxygenases for Biosynthesis of Actinorhodin and Related
Antibiotics
Authors: Koji Ichinose, Makoto Hashimoto, Takaaki Taguchi, Kazuki
Ishikawa, Ryuichiro Mori, Akari Hotta, Susumu Watari,
Kazuaki Katakawa, Takuya Kumamoto, and Susumu
Okamoto
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemBioChem 10.1002/cbic.201900490
Link to VoR: http://dx.doi.org/10.1002/cbic.201900490
A Journal of
www.chembiochem.org

10.1002/cbic.201900490
ChemBioChem
COMMUNICATION
Unveiling Two Consecutive Hydroxylations: Mechanism of
Aromatic Hydroxylations Catalyzed by Flavin-Dependent
Monooxygenases for Biosynthesis of Actinorhodin and Related
Antibiotics
Makoto Hashimoto†[a], Takaaki Taguchi†[b], Kazuki Ishikawa[a], Ryuichiro Mori[a], Akari Hotta[a], Susumu
Watari[a], Kazuaki Katakawa[a], Takuya Kumamoto[c], Susumu Okamoto[d] and Koji Ichinose*[a]
Abstract: Flavin-dependent monooxygenases are ubiquitous in living
systems and are classified into single- or two-component systems.
Actinorhodin produced by Streptomyces coelicolor is a representative
polycyclic polyketide that is hydroxylated by the ActVA-5/ActVB two-
component hydroxylase system. These homologous systems are
widely distributed in bacteria, but their reaction mechanisms remain
unclear. The present in vitro investigation provided chemical proof of
N-hydroxylase for auxin biosynthesis,[2] whereas the yeast yFMO
is involved in disulfide bond formation during protein folding
through regulation of the glutathione/glutathione disulfide ratio.[3]
A plethora of microbial FMOs have been reported to be involved
in the degradation of environmental aromatic compounds, and in
the biosynthesis of secondary metabolites.[4] These FMOs are
classified into two types: single-component systems have both
flavin reduction and oxygenation domains in a single polypeptide,
whereas two-component systems comprise distinct flavin
reductase and oxygenase proteins. Functional and structural
studies of FMOs have mostly focused on the single-component
systems. Consequently, few two-component systems have been
characterized at the mechanistic level, particularly those for the
biosynthesis of complex natural products.[5, 6]
two
consecutive
hydroxylations
via
hydroxynaphthalene
intermediates involved in actinorhodin biosynthesis. The ActVA-5
oxygenase component catalyzed a stepwise dihydroxylation of the
substrate, whereas the ActVB flavin reductase not only supplied a
reduced cofactor, but also regulated the quinone-hydroquinone
interconversion of an intermediate. Our study provides clues for
understanding the general biosynthetic mechanism of highly
functionalized aromatic natural products with structural diversity.
Here, we focus on the pyranonaphthoquinone antibiotic
actinorhodin (ACT, 1) produced by Streptomyces coelicolor A3(2)
as the sole known example of a polycyclic aromatic compound
hydroxylated by a two-component FMO (Scheme 1). The FMO for
ACT biosynthesis is the ActVA-ORF5/ActVB system consisting of
an oxygenase, ActVA-5, and a flavin:NADH oxidoreductase,
ActVB. Previously, we demonstrated in vivo that the system was
a bifunctional oxygenase, governing quinone-formation at C-6
and hydroxylation at C-8 of a tricyclic ACT intermediate, 6-deoxy-
dihydrokalafungin (DDHK, 2) (Scheme 1).[7] Although 2 was
deduced to be the natural substrate of the system,[7] it has never
been isolated from any biosynthetic mutants or strains of S.
coelicolor due to an oxidative dimerization leading to the formation
Flavin-dependent monooxygenases (FMOs) are distributed in
animals, plants, and microorganisms, where they play essential
roles in a variety of biological processes. Animal FMOs primarily
oxygenate heteroatom-containing drugs and xenobiotics,
producing metabolites such as N- and S-oxides that are readily
excreted from the body.[1] The plant FMO YUCCA is a tryptamine
[a]
Dr. M. Hashimoto[†], Dr. K. Ishikawa, R. Mori, A. Hotta, S. Watari,
Dr. K. Katakawa, Prof. Dr. K. Ichinose
Research Institute of Pharmaceutical Sciences, Musashino
University
of
a
shunt product, actinoperylone.[8] Thus, in vitro
1-1-20, Shinmachi, Nishitokyo-shi, Tokyo 202-8585, Japan
E-mail: ichinose@musashino-u.ac.jp
Dr. T. Taguchi[†]
National Institute of Health Sciences
3-25-26, Tonomachi, Kawasaki-ku, Kawasaki-shi, Kanagawa 210-
9501, Japan
characterizations were limited by this fact, and emodinanthrone
was used as a surrogate tricyclic substrate.[7]
[b]
[c]
The present study describes a semisynthetic preparation of 2
from (S)-DNPA (3), an isolable intermediate of ACT biosynthesis
(Scheme 1), and the unambiguous chemical proof of the two
consecutive hydroxylations of 2, revealing the previously
uncharacterized machinery for the interconversion between the
quinone- and hydroquinone-form of intermediates governed by
the ActVA-5/ActVB system. Furthermore, we demonstrated the
monohydroxylation or dihydroxylation activity of the closest
homologs of ActVA-5, Med-7 and Gra-21, for medermycin (MED,
4) and granaticin (GRA, 5) biosyntheses (Scheme 1). We propose
that iterative oxidation catalyzed by two-component FMOs
provides an explanation for the structural diversity in oxygen-
containing natural products.
Prof. Dr. T. Kumamoto
Graduate School of Biomedical and Health Sciences, Hiroshima
University
1-2-3 Kasumi, Minami-ku, Hiroshima City, Hiroshima 734-8553,
Japan
[d]
[†]
Dr. S. Okamoto
National Agriculture and Food Research Organization
2-1-12 Kannondai, Tsukuba, Ibaraki 305-8642, Japan
These authors contributed equally to this work
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900490
ChemBioChem
COMMUNICATION
slightly acidic conditions (pH 6.5) and methanol as a solubilizing
agent for 2, followed by a mild heat treatment (see Supporting
Information) for enzyme inactivation prior to HPLC analysis. The
resultant assay workflow produced about 23% loss of 2 in the
absence of ActVA-5/ActVB (Figure S14). After enzymatic reaction,
we detected the peaks of hydroxy-tetrahydrokalafungin (THK-OH,
8) and 8-hydroxy-dihydrokalafungin (DHK-OH, 9), both of which
were previously isolated from a deletion mutant of the actVA-
ORF4 gene of S. coelicolor A3(2).[11] Their identical RT and UV-
VIS spectra with the authentic ones were confirmed (Figure 1A,
Figure S15), and notably, the conversion of 2 to 8 was almost
quantitative (93 % yield). Without ActVA-5 or VB, unreacted 2
together with a small amount of 7 was detected (Figure 1B, 1C);
7 is the quinone form of a postulated C-6 oxygenated product,
T3HN (10), possibly derived from autoxidation. Additionally, a
series of control experiments were conducted to evaluate the
requirement of each component in the assay mixture (Figure
S16): 1) a higher amount of 2 was detected in the absence of
enzymes; 2) the spontaneous formation of 7 was up to 10 % that
of 2; 3) NADH and FMN were essential for the activity; 4) the
omission of catalase led to a reduction in activity. The combined
results clearly showed that 2 is indeed the native substrate of the
ActVA-5/ActVB system to conduct the two consecutive
hydroxylations at C-6 and C-8.
When 7 was subjected to the assay, the formation of 8 was
observed, suggesting that 7 is once reduced back to 10 by the
ActVA-5/ActVB system (Figure 1D). The C-8 hydroxylation of 7
was not detected in our previous investigation,[7] where ethylene
glycol monomethyl ether (EGME) was included in the assay
mixture as a solubilizing agent. This result might have been due
to the presence of hydrogen peroxide, an inevitable byproduct in
the assay mixture, which is known to convert EGME to 2-
methoxyacetoaldehyde, which might inhibit the conversion of 7 to
10.[13]
Scheme 1. The proposed biosynthetic pathways of 1, 4 and 5. Position number
is based on the polyketide (C16) biosynthetic origin. A common bicyclic
intermediate[9] undergoes stereospecific keto-reductions at C-3/C-15 leading to
either 2 or 12. Two hydroxylations at C-6/C-8 and a single hydroxylation at C-6
are involved in ACT/GRA and MED biosynthesis, respectively.[10] ActVA-4 is a
postulated enzyme involved in a dimerization step during ACT biosynthesis.[11]
The dotted arrows depict chemical conversions of the related metabolites.
The successful second hydroxylation of 7 at C-8 resulted in
notable HPLC peaks X and Y at ca. 3 min (Figure 1D).
Considering the co-presence of the quinone-form of the C-6/C-8
oxygenated products, 7 and 9, we assumed X and Y to be the
immediate enzymatic products, 10 and T4HN (11), respectively.
The treatment of 7 solely with ActVB produced a substantial
amount of X (3.2 min), whose formation was confirmed to be
dependent on ActVB, NADH, and FMN (Figure 1E, S17). The
similarity in UV-VIS spectra of X and 2 also suggested their
naphthalene chromophores (Figure S18). LC/HRESIMS analysis
of the reaction mixture containing X revealed a molecular formula
of C16H16O6 exactly matching that of 10 (Figure S18). An earlier
study showed that sodium dithionite (Na2S2O4) converts an
[12]
For preparation of substrate 2, 3
was reduced by NaBH4,
resulting in two compounds, which were assumed to be 2 or its C-
15 epimer, epi-DDHK (6), from UV-VIS and MS spectra (RT 19
min and 20 min, Figure S1). Each was purified by preparative
HPLC, and subjected to extensive NMR spectral analysis (Figure
S1-S13, Table S1, S2). These spectra established that their
planar structure is 2. The NOESY spectrum of the compound with
the earlier HPLC retention time (Figure S7) indicated an NOE
correlation between 3-H and 16-H, proving its structure as 2 with
the (3S, 15R) configurations. Likewise, the reasonable NOE
correlation for the other compound (Figure S13) elucidated its
absolute structure as 6. Thus, we succeeded in establishing the
semisynthetic method for preparing 2 together with 6 for the first
time. Since unstable properties prone to decomposition of 2 were
observed, it was stored at -30°C under a nitrogen atmosphere.
First, we optimized in vitro assay conditions for ActVA-5/ActVB
based on our previous studies with model substrates.[7] To
minimize an unwanted degradation of 2, we employed both
anthraquinone to the corresponding anthrahydroquinone[14]
,
which led us to treat 7 with Na2S2O4 to demonstrate the chemical
formation of X (Figure S18). In fact, peak X was present in the
HPLC analysis of the standard assay mixture (Figure 1A), most
likely to be undergoing the second hydroxylation. All the results
strongly suggest X to be 10, itself.
Similar ActVB-dependent formation of Y (3.3 min) from 9 was
observed (Figure 1F). Its rigorous LC/HRESIMS analysis resulted
in the MS detection corresponding to a molecule of C16H16O8
(Figure S20), which is an oxygenated formula of 11 or 8. Although
Na2S2O4 treatment readily converted 9 to 8, the reduction in the
amount of Na2S2O4 selectively produced Y instead of 8 (Figure S
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900490
ChemBioChem
COMMUNICATION
21). In fact, Y was detectable under the standard assay conditions
(Figure 1A, 1D), indicating that Y could be an oxygenated shunt
product derived from 11 or 8.
The conversion by ActVB from the quinone (7 and 9) to
hydroquinone (10 and 11) form of substrates is an additional
important observation in this study. This in vitro quinone-reduction
activity of ActVB was previously reported under anaerobic
conditions[15] and it is now extended to aerobic conditions,
possibly reflecting the significance of a separate reductase for an
adequate supply of both reduced flavin cofactors and substrates.
Time-course analysis of 2, 7, and 8 under the standard assay
condition revealed that 2 rapidly decreased within 1 min, and 8
was generated at 63% yield., while almost no 7 was detected
(Figure 2A). Subsequent analysis with a reduced amount of
ActVA-5 over 20 min revealed the accumulation of 7 in
association with the decrease in 2, followed by the significant
generation of 8 (Figure 2B), favoring the step-wise recognition by
the ActVA-5. Our earlier study[11] on an actVA-ORF4 deletion
mutant (Scheme 1) suggested its involvement in the dimerization
step of the ACT biosynthesis. This is based on the accumulation
of 9 as well as 7 in the mutant, being in agreement with the step-
wise hydroxylation steps proposed in this study.
Figure 1. HPLC chromatograms of the reaction mixtures of various substrates
with combinations of ActVA-5 and ActVB: (A) 2 with ActVA-5/ActVB; (B) 2 with
ActVB; (C) 2 with ActVA-5; (D) 7 with ActVA-5/ActVB; (E) 7 with ActVB; (F); 9
with ActVB. See Supporting Information for details of the reaction conditions.
Figure 2. Time-course analysis of the ActVA-5/ActVB system: (A) the standard
assay; (B) the assay with a reduced concentration (one-fifth) of ActVA-5 (means
±S. E., n = 3)
The mechanism of two consecutive hydroxylations is of great
interest. Two mechanisms could be assumed: 1) once ActVA-5
accepts 2, it catalyzes two hydroxylations at C-6 and C-8
consecutively in the active site(s) and releases the product 11
(single substrate recognition), or 2) ActVA-5 catalyzes the first
hydroxylation at C-6 and releases the product 10, which was 8-
Our previous in vivo analysis showed that two ActVA-5 homologs,
Med-7 and Gra-21 for the biosynthesis of 4 and 5 (Scheme 1),
also have DDHK hydroxylation activity.[7] Their in vitro activities
were also investigated using 2 and ActVB as a flavin reductase
(Figure 3). The conversion yields of 2 to 8 and 9 of ActVA-5 and
Gra-21 were 99% and 80%, respectively. Moreover, Gra-21 was
hydroxylated by
a separate ActVA-5 (stepwise substrate
recognition). In both cases, two molecules of C(4a)-FMN-
hydroperoxide species (FlHOO-)[16] are required in the reaction
stoichiometry. The C2 component of 4-hydroxyphenylacetate 3-
hydroxylase from Acinetobacter baumannii shows 44% similarity
to ActVA-5 and its X-ray crystallographic study indicated that the
C2 protein accommodates a single molecule of FMN in its catalytic
cavity.[17] Modeling studies on the ActVA-5 protein using C2
protein (PDB ID: 2JBT) as a template (Figure S22, S23) allowed
us to propose that the protein cavity would accommodate a single
molecule of FMN.
confirmed to be
a bifunctional FMO for C-6 and C-8
hydroxylations. Interestingly, the native substrate of Gra-21
should be enantio-DDHK (12) with (3R, 15S) configurations
(Scheme 1). The stereochemical centers at C-3 and C-15 exerted
no major influence on the C-6 and C-8 hydroxylation activities of
Gra-21.
Med-7 has been suggested to be a monofunctional FMO.[7]
Indeed, 7 was the main product of the in vitro hydroxylation
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900490
ChemBioChem
COMMUNICATION
reaction, and the conversion yield of 2 to 7 was 45%. Interestingly,
a small amount of 8, equivalent to 6% of the substrate, was
detected. These observations lead to the hypothesis that the
affinity of Med-7 for 10 could be lower than those of ActVA-5 and
Gra-21. Thus, Med-7 releases 10 immediately after C-6
hydroxylation, and the second hydroxylation reaction does not
take place.
coupling, the last step in ACT biosynthesis. Recently, we
established an in vitro reconstitution system using recombinant
ACT biosynthetic enzymes to produce 3.[9] In S. coelicolor, an
enoylreductase encoded by actVI-ORF2 converts
3
stereospecifically to 2 (Scheme 1). [7, 8] Our ongoing studies are
dealing with the functional expression of ActVI-2 to connect the
earlier pathway to the present ActVA-5/ActVB system to
reconstitute at least 25 steps from malonyl-CoA, covering the
entire ACT biosynthetic pathway except for dimerization.
These functional differences between ActVA-5 and Med-7
should also be derived from the distinct structural determinants of
substrate recognition. Modeling studies were extended to Gra-21
and Med-7 (Figure S22, 23), followed by the docking simulation
of 2 and 10. A marked difference was observed for the residues
in the vicinity of ligand pockets (Figure S24). The sets of two
aligned residues (227R and 314R/ActVA-5; 255K and 336R/Gra-21;
227K and 313R/Med-7) are particularly of interest. The two arginine
residues of ActVA-5 (Figure S25), which function as a proton-
donor to an anionic form of 2, tightly accommodate a substrate.
The same role would be undertaken by the single residue:
255K/Gra-21 (Figure S26) and 358S/Med-7 (Figure S27). It is
noteworthy that 358S/Med-7 is distantly positioned from
227R/ActVA-5 and 255K/Gra-21 in both the alignment and the
models of the three related proteins (Figure S24). The Med-7
structure seems to form an apparently looser catalytic pocket for
2 (Figure S27) and 10 (Figure S28) than those of ActVA-5 (Figure
S25) and Gra-21 (Figure S26). This difference could be explained
by the lack of pi-stacking (234F/ActVA-5, 262F/Gra-21, 234L/Med-7)
Two-component FMOs homologous to the ActVA-5/ActVB
system are widely distributed in bacteria (Figure S29, Table S3,
S4). This work demonstrated that a flavin reductase, ActVB,
apparently regulates the interconversion between the
naphthoquinone and hydronaphthoquinone form of ACT
biosynthetic intermediates. This observation agrees with the
recent finding of the importance of hydronaphthoquinone in
biosynthesis,[18] and our results would provide an example of the
general biosynthetic mechanism of highly functionalized aromatic
natural products such as polyphenols and polymeric aryl
compounds.
In conclusion, the ActVA-5/ActVB system provides a previously
uncharacterized example of the hydroxylation of a polycyclic
aromatic compound catalyzed by a two-component FMO.
and hydrophobic
(
122M/ActVA-5, 147M/Gra-21, 120A/Med-7)
interactions (Figure S25-S27) with a ligand molecule, which would
contribute to the difference in turnover of a substrate and/or FNM,
accounting for the mono-functionality of Med-7.
Acknowledgements
We thank Mr. Y. Kanai, Mr. K. Maru, and Ms. E. Hoshi, Faculty of
Pharmacy, Musashino University for their assitance in
experiments. We also thank Mr. Yoshirou Kimura, Molsis Inc. for
helpful discussion. This work was supprted by JSPS KAKENHI
Grant Number JP17K08350 (to TT) and JP19H03385 (to KI).
Cnflict of Interest
The authors declare no conflict of interest.
Keywords: aromatic hydroxylation• biosynthesis• flavin-
dependent monooxygenase• actinorhodin • reaction mechanisms
[1]
[2]
S.K. Krueger, D.E. Williams, Pharmacol. Ther., 2005, 106, 357-387.
Y. Zhao, S.K. Christensen, C.Fankhauser, J.R. Cashman, J.D. Cohen,
D. Weigel, J. Chory, Science, 2001, 291, 306-309.
[3]
[4]
[5]
[6]
[7]
J.-K. Suh, L.L. Poulsen, D.M. Ziegler, J.D. Robertus, Proc. Natl. Acad.
Sci. USA, 1999, 96, 2687-2691.
Figure 3. THK-OH production activities of ActVA-5, Med-7, and Gra-21 from 2
with ActVB as a FMN reductase (means ±S. E., n = 3)
M.M.E. Huijbers , S. Montersino, A.H. Westphal, D. Tischler, W.J.H. van
Berkel, Arch. Biochem. Biophys., 2014, 544, 2-17.
T. Heine, W.J.H. van Berkel, G. Gassner, K.-H.van Pée, D. Tischler,
Biology 2018, 7, 42; doi:10.3390/biology7030042
We successfully established a semisynthetic procedure for
preparing 2 and 6 from a biosynthetic intermediate 3 and an assay
condition to achieve two consecutive hydroxylations at C-6 and
C-8 of 2 to produce 8. The bifunctionality of ActVA-5 indicated by
in vivo analysis[7] was unambiguously proved by in vitro analysis.
Time-course analysis suggests that the stepwise recognition
mechanism would be supported. The fact that 2 is converted to 8
by ActVA-5/ActVB supports our previous proposal[11] that 11
would be a substrate for dimerization by regiospecific aryl-
V. Weichold, D. Milbredt, K.-H. van Pee, Angewandte Chemie Int. Ed.,
2016, 55, 6374-6389.
a) T. Taguchi, M. Yabe, H. Odaki, M. Shinozaki, M. Metsä-Ketelä, T. Arai,
S. Okamoto, K. Ichinose, Chem. Biol., 2013, 20, 510-520; b) S. Okamoto,
T. Taguchi, K. Ochi, K. Ichinose, Chem. Biol., 2009, 16, 226-236.
T. Taguchi, S. Okamoto, T. Itoh, Y. Ebizuka, K. Ochi, K. Ichinose,
Tetrahedron Lett., 2008, 49, 1208-1211.
[8]
[9]
T. Taguchi, T. Awakawa, Y. Nishihara, M. Kawamura, Y. Ohnishi, K.
Ichinose, ChemBioChem., 2017, 18, 316-323.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900490
ChemBioChem
COMMUNICATION
[10] a) K. Ichinose, D.J. Bedford, D. Tornus, A. Bechthold, M.J. Bibb, W.P.
Revill., H.G. Floss, D.A. Hopwood, Chem. Biol., 1998, 5, 647-659; b) K.
Ichinose, M. Ozawa, K. Itou, K. Kunieda, Y. Ebizuka, Microbiology, 2003,
149, 1633-1645.
[15] a) J. Valton, M. Fontecave, T. Douki, S.G. Kendrew, V. Nivière, J. Biol.
Chem., 2006, 281, 27-35; b) J. Valton, C. Mathevon, M. Fontecave, V.
Nivière, D.P. Ballou, J. Biol. Chem., 2008, 283, 10287-10296.
[16] K. Thotsaporn, J. Sucharitakul, J. Wongratana, C. Suadee, P. Chaiyen,
Biochim. Biophys. Acta, 2004, 1680, 60-66.
[11] T. Taguchi, T. Ebihara, A. Furukakwa, Y. Hidaka, R. Ariga, S. Okamoto,
K. Ichinose, Bioorg. Med. Chem. Lett., 2012, 22, 5041-5045.
[12] T. Taguchi, Y. Ebizuka, D.A. Hopwood, K. Ichinose, J. Am. Chem. Soc.,
2001, 123, 11376-11380.
[17] A. Alfieri, F. Fersini, N. Ruangchan, M. Prongjit, P. Chaiyen, A. Mattevi,
Proc. Natl. Acad. Sci. USA, 2007, 104, 1177-1182.
[18] S.M. Husain, M.A. Schäzle, S. Lüdeke, M. Müller, Angewandte Chemie
Int. Ed., 2014, 53, 9806-9811.
[13] A. Fisher, C. Hahn, Water Res., 2005, 39, 2002-2007.
[14] M.A. Schäzle, S.M. Husain, S. Ferlaino, M. Müller, J. Am. Chem. Soc.,
2012, 134, 14742-14745.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900490
ChemBioChem
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 1:
COMMUNICATION
Filling the gap: 6-Deoxy-
dihydrokalafungin (DDHK) is a
previously uncharacterized
Makoto Hashimoto†, Takaaki Taguchi†,
Kazuki Ishikawa, Ryuichiro Mori, Akari
Hotta, Susumu Watari, Kazuaki
intermediate for actinorhodin (ACT)
biosynthesis. Semisynthetic
Katakawa, Takuya Kumamoto, Susumu
Okamoto and Koji Ichinose*
preparation of DDHK was
Page No. – Page No.
established, followed by its use for in
vitro enzymatic reactions using a
flavin-dependent monooxygenase,
ActVA-ORF5, and a flavin reductase,
ActVB. DDHK was hydroxylated
stepwise at C-6 and C-8, proving its
ACT biosynthetic intermediacy.
Unveiling Two Consecutive
hydroxylations: Mechanism for
Aromatic Hydroxylations Catalyzed
by Flavin-Dependent
Monooxygenases for Biosynthesis of
Actinorhodin and Related Antibiotics
This article is protected by copyright. All rights reserved.