Characterization of 2-Oxindole Forming Heme Enzyme MarE, Expanding the Functional Diversity of the Tryptophan Dioxygenase Superfamily

Article abstract of DOI:10.1021/jacs.7b05517

3-Substituted 2-oxindoles are important structural motifs found in many biologically active natural products and pharmaceutical lead compounds. Here, we report an enzymatic formation of the 3-substituted 2-oxindoles catalyzed by MarE in the maremycin biosynthetic pathway in Streptomyces sp. B9173. MarE is a homologue of Fe^II/heme-dependent tryptophan 2,3-dioxygenases (TDOs). Typical TDOs usually catalyze the insertion of two oxygen atoms from O₂ into an indole ring to generate N-formylkynurenine (NFK)-like products. In contrast, MarE catalyzes the insertion of a single oxygen atom from O₂ into an indole ring, to probably generate an epoxyindole intermediate that undergoes an unprecedented 2,3-hydride migration to form 2-oxindole structure. MarE shows substrate robustness to catalyze the conversion of a series of 3-substituted indoles into their corresponding 3-substituted 2-oxindoles. Although containing most key amino acid residues conserved in well-known TDO homologues, MarE falls into a separate new subgroup in the phylogenetic tree. The characterization of MarE and its homologue enriches the functional diversities of TDO superfamily and provides a new strategy for discovering novel natural products containing 3-substituted 2-oxindole pharmacophores by genome mining.

Full text of DOI:10.1021/jacs.7b05517

Article
pubs.acs.org/JACS
Characterization of 2‑Oxindole Forming Heme Enzyme MarE,
Expanding the Functional Diversity of the Tryptophan Dioxygenase
Superfamily
Yuyang Zhang,† Yi Zou,†,‡ Nelson L. Brock,† Tingting Huang,† Yingxia Lan,† Xiaozheng Wang,†
Zixin Deng,† Yi Tang,‡ and Shuangjun Lin*,†
†State Key Laboratory of Microbial Metabolism, Joint International Laboratory on Metabolic and Developmental Sciences, School of
Life Sciences and Biotechnology, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai 200240, China
‡Department of Chemical and Biomolecular Engineering, Department of Chemistry and Biochemistry, and Department of
Bioengineering, University of California, Los Angeles, 5531 Boelter Hall, 420 Westwood Plaza, Los Angeles, California 90095, United
States
S
* Supporting Information
ABSTRACT: 3-Substituted 2-oxindoles are important struc-
tural motifs found in many biologically active natural products
and pharmaceutical lead compounds. Here, we report an
enzymatic formation of the 3-substituted 2-oxindoles catalyzed
by MarE in the maremycin biosynthetic pathway in
Streptomyces sp. B9173. MarE is a homologue of FeII/heme-
dependent tryptophan 2,3-dioxygenases (TDOs). Typical
TDOs usually catalyze the insertion of two oxygen atoms
from O2 into an indole ring to generate N-formylkynurenine
(NFK)-like products. In contrast, MarE catalyzes the insertion of a single oxygen atom from O2 into an indole ring, to probably
generate an epoxyindole intermediate that undergoes an unprecedented 2,3-hydride migration to form 2-oxindole structure.
MarE shows substrate robustness to catalyze the conversion of a series of 3-substituted indoles into their corresponding 3-
substituted 2-oxindoles. Although containing most key amino acid residues conserved in well-known TDO homologues, MarE
falls into a separate new subgroup in the phylogenetic tree. The characterization of MarE and its homologue enriches the
functional diversities of TDO superfamily and provides a new strategy for discovering novel natural products containing 3-
substituted 2-oxindole pharmacophores by genome mining.
the indole ring was found essential for the rearrangement via
the pseudo-para-quinone methide formation. In the case where
the indole ring lacks this kind of oxygen substituent, the 2-
oxindole formation can then be catalyzed by the second type
enzymes known as cytochrome P450 monooxygenases, such as
FtmG from the biosynthetic pathway of spirotryprostatin B.5
FtmG was proposed to catalyze two rounds of hydroxylation,
which were followed by the semipinacol-type rearrangement
without an epoxide intermediate.5
The 2-oxindole moiety was present in maremycins A−D (3-
substituted 3-hydroxy-2-oxindole) and maremycins E−G
(spirocyclic 2-oxindole) (Figure 1A).6 Surprisingly, when
examining the biosynthetic gene cluster (mar) that is
responsible for the biosynthesis of maremycin A−G from
Streptomyces sp. B9173, neither homologue of the FMO nor
P450 can be found among the 17 mar open reading frames
(Figure 2A).6c,7 We previously showed that the three gene
cassette of marG (encoding aminotransferase), marH (encod-
ing isomerase), and marI (encoding methyltransferase) was
INTRODUCTION
■
Natural products containing the 2-oxindole structure, which
can be either a 3-substituted 3-hydroxy-2-oxindole or a 2-
oxindole spirocycle, possess a wide spectrum of biological
activities, including anticancer, inhibition of tubulin polymer-
ization, neuroprotection, and antihypertension, etc.1 (Figures 1
and S1). This pharmacophore is also extensively incorporated
into synthetic drug candidates, including NITD609, SM-
130686, and SAR405838 (Figures 1 and S2)2 as well as
marketed anticancer drugs Sunitinib and Palladia.3 Discovering
enzymes that can generate the 2-oxindole structure under mild
conditions can therefore be useful for building block synthesis
and late-stage transformations.
Two types of enzymes have been reported to catalyze
formation of 2-oxindole moiety of nearly mature natural
products. The first type is flavin-dependent monooxygenases
(FMOs), such as NotB that catalyzes the conversion of
notoamide E to notoamide C as a minor product4 and FqzB
that catalyzes the biosynthesis of spirotryprostatin A.5 FMOs
were proposed to catalyze the formation of 2-oxindoles through
an epoxide intermediate and a subsequent semipinacol-type
rearrangement.4,5 For FMOs, the oxygen substituent at C6 of
Received: May 28, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.7b05517
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
of 2-oxindole was proposed to undergo a unique 2,3-hydride
migration involving an epoxyindole intermediate. The
phylogenetic analysis showed that MarE is located in a separate
new subgroup of TDO superfamily, although it contained most
of the conserved key amino acid residues. This work not only
expands the functional diversity of tryptophan dioxygenase
superfamily but also provides a new strategy to discover active
natural products containing 3-substituted 2-oxindole motifs by
genome mining methods.
RESULTS
■
In Vivo Characterization of MarE. To investigate if MarE
was functional in the biosynthetic pathway of maremycins,
marE gene was inactivated in Streptomyces sp. B9173 with the
insertion of aac(3)IV (apramycin resistance gene) by λ-red
recombination system (Figure S3). The resulting mutant
ΔmarE abolished the production of maremycins and did not
accumulate any detectable biosynthetic intermediates (or shunt
products). The production could be restored to wild-type levels
by genetic complementation with the marE gene (Figure 2C).
These results suggested that marE was indeed directly related
to the production of maremycins, but the oxidation of indole
ring catalyzed by MarE may take place at the precursor stage
instead of on a more mature intermediate.
Biochemical Characterization of MarE and Identifica-
tion of the Product. For in vitro studies, MarE was
overexpressed and purified to homogeneity with an N-terminal
his6 tag from Escherichia coli BL21 (DE3) (Figure S4). Purified
MarE protein displayed a brownish red color (Figure S4) and
the characteristic ultraviolet−visible (UV−vis) absorbance of
heme-containing proteins with a Soret band at 405 nm (Figure
2D). Size-exclusion chromatography revealed that MarE
formed a homotetramer with the size of about 130 kDa similar
to the characterized TDOs10,11 (Figure S5).
Figure 1. Natural products and synthetic leads containing 2-oxindole.
Both 3-hydroxy-2-oxindole and spirocyclic 2-oxindole are shown in
red.
responsible for the formation of the maremycin building block
(2′S,3′S)-methyltryptophan (1),7 starting from tryptophan
(Figure 2B). MarF is also found in the same operon, which
catalyzes N-methylation of the indole ring at the later stage
than 2-oxindole formation,6c and MarE, which is a homologue
of the well-studied tryptophan 2,3-dioxygenase (TDO) super-
family. TDOs are heme-containing oxygenases that are found in
nearly all organisms and have crucial roles in the oxidative
cleavage of tryptophan into N-formylkynurenine (NFK),8
which is the first and rate-limiting step in the tryptophan
degradation pathway. In physiological conditions, TDO
systematically regulates the tryptophan levels, and as such,
TDO is becoming a novel and promising therapeutic target in
the clinical disease treatment in recent studies,9a,b analogous to
Homo sapiens indolamine 2,3-dioxygenase 1 (HsIDO1) which
was well-known as an immune system suppression target.9c
Although TDO acts typically as a dioxygenase that inserts both
oxygen atoms into the cleaved product, we hypothesize that
MarE may be responsible for formation of 2-oxindole in the
maremycin biosynthetic pathway. To the best of our knowl-
edge, no TDO catalyzing the formation of the 2-oxindole by
insertion of single oxygen atom to the product, like a
monooxygenase, has been identified to date.
Here, the role of MarE in the biosynthetic pathway of
maremycins from Streptomyces sp. B9173 was investigated by in
vivo genetic experiments and in vitro biochemical assays.
Although belonging to the TDO superfamily, MarE was
biochemically characterized to catalyze the insertion of a single
oxygen atom from O2 into a series of 3-substituted indoles to
generate 3-substituted 2-oxindoles, instead of the insertion of
two atoms of oxygen to form C−C double bond cleaved
products catalyzed by canonical TDOs. By using the isotope-
labeled substrate, the mechanism of MarE-catalyzed formation
Our previous work showed that MarG, MarH, and MarI
catalyzed the net C-methylation of L-tryptophan to yield 1
(Figure 2B).7 To examine if MarE reacts immediately
downstream using 1 as a substrate, we first incubated MarE
with 1, which induced a shift of the Soret band from 405 to 408
nm without accumulation of oxidation products (Figure 2D and
2E). The shift of the Soret band is highly consistent with that
found for CmTDO from Cupriavidus metallidurans12 and
HsTDO from Homo sapiens.13 The shift of the Soret band
also well matches TDOs featuring an induced-fit model,10,11
indicating that 1 may enter the active site of enzyme and bind
to MarE. In general, the reactions catalyzed by TDOs were
dependent on the iron at the ferrous (FeII) state, although
HsTDO and TDO from mosquito were reported to have
activity at their ferric (FeIII) state.13 When ascorbate, commonly
used as a reducing agent for assaying TDOs,14 was incubated
with MarE, the Soret band of MarE at 405 nm decreased and an
absorption shoulder at 425 nm developed, which agrees with
the features of TDOs at the ferrous (FeII) state12 (Figure 2D).
Consequently, when both 1 and ascorbate were incubated with
MarE, a new product 2 appeared (Figure 2E) and the Soret
band of MarE at 408 nm decreased, while the Soret band at 425
nm increased (Figure 2D). The high-resolution electrospray
ionization mass spectrometry (HRESIMS) analysis of the
product 2 showed an ion peak at m/z 235.1087 (calcd
+
235.1077 for C12H15N2O3 ), 16 Da greater than that of 1
(Figure S6). The mass of 2 suggests that at the ferrous (FeII)
state of MarE, only a single atom of oxygen was inserted into
B
DOI: 10.1021/jacs.7b05517
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 2. In vivo and in vitro characterization of MarE. (A) The marE gene locus in biosynthetic gene cluster of maremycins in Streptomyces sp.
B9173 (the green dashed box is a predicted operon including the five gene cassette marEFGHI). (B) The predicted sequential biosynthetic route of
precursor for maremycins. (C) In vivo characterization (detection at 254 nm). (i) Maremycin A authentic standard; (ii) Streptomyces sp. B9173 wild-
type strain (★, Maremycin A; ☆, Maremycin B); (iii) ΔmarE mutant strain; (iv) ΔmarE mutant strain complemented with additional marE gene.
(D) The UV−vis spectra of MarE at different conditions. (i) MarE (blue line) and blank control (Buffer C, red line). (ii) MarE (blue line) and MarE
with 1 mM 1 mixture (red line). (iii) MarE (blue line) and MarE with 10 mM ascorbate mixture (red line). (iv) MarE (Blue line) and MarE with 10
mM ascorbate and 1 mM 1 mixture (Red line). (All of the insets were partially enlarged for each figure.) (E) In vitro characterization (detection at
254 nm). (i) 1; (ii) 1 incubated with MarE; (iii) 1 incubated with MarE and ascorbate for 12 h; (iv) 1 incubated with MarE and ascorbate under
anaerobic conditions.
the substrate, in contrast to the canonical insertion of two
oxygen atoms catalyzed by known TDOs.
To identify the structure of the MarE-catalyzed monooxyge-
The Effect of Reducing Agents on the Activity of
MarE. In general, heme-dependent dioxygenases (TDOs and
IDOs) require the reducing agents to keep the bound iron in
enzymes at the ferrous (FeII) state for activation of O2 in the
beginning of their catalytic cycles in the in vitro assays.
Although ascorbate was commonly used as a reducing agent in
the TDO assays,14 the small UV−vis absorption shoulder
emerged at 425 nm after adding 10 mM ascorbate (Figure 2D),
indicating that a small fraction of enzymes at the FeII state was
formed. Glutathione (GSH), an important physiological
reducing agent, has a higher reducing potential than
ascorbate.15 To rule out the possibility that the reducing
potential of ascorbate was not sufficient for dioxygenase
activity, glutathione was tested in the MarE-catalyzed reaction
system. However, although MarE was indeed almost
completely converted to the ferrous (FeII) state (Figure S10),
no dioxygenase activity can be detected (Figure S11). In the
presence of glutathione, we observed a Soret maximum at 423
nm and two Q-bands in the 500−600 nm region, suggestive of
a hexa-coordinating heme species that suggests glutathione
nation product 2, the reaction was performed on a preparative
1
scale, and the product was purified for NMR analysis. Its H
and 13C NMR data (Table S1, Figures S7 and S8) are very
similar to those reported for 1.7 The only observed differences
are that 2 misses an aromatic proton and two aromatic carbons,
and instead a carbonyl carbon signal occurred at δC 181.9 ppm
in 2. The 2D NMR spectra, especially the heteronuclear
multiple bond correlation (HMBC), revealed that a proton
signal and a carbon signal overlapped with those of methanol-d4
(Figure S8). Taken together, the product 2 was verified as 3-
substitued 2-oxindole. When the MarE assay was carried out for
longer time, no dioxygenase activity, that is, the cleavage of 1
into the corresponding 3-methyl-NKF could be detected at the
experimental condition (Figure S9). Therefore, MarE was
confirmed as the oxygenase to catalyze the formation of the 2-
oxindole building block in maremycins, thus adding a new
function to the TDO family.
C
DOI: 10.1021/jacs.7b05517
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
coordination.13a This suggests that glutathione may indeed be
an inhibitor in this reaction and explains the lowered activity
observed.16 We also tested inorganic reductant dithionite, but
similar results were found (Figure S12). These data
demonstrate that the reductants have effects on MarE activity,
but it remains unclear what is the physiological reductant of
MarE in this reaction.
Substrate Specificity of MarE. To test the substrate
promiscuity of MarE and its potential utility as a biocatalyst, a
series of 1 analogs including indole (1b), 3-methylindole (1c),
L-tryptophan (1d), D-tryptophan (1e), tryptamine (1f), indole-
3-propionic acid (1g), (2′S, 3′R)-methyltryptophan (1h), 1-
methyl-L-tryptophan (1i), and 1-methyl-indole (1j) were tested
as putative substrates (Figures 3A and S13). High-resolution
dioxygen were sequentially, not simultaneously, inserted into
the substrate with an unstable epoxide as the key intermediate
in the reactions (Scheme S1).14,17,18 Interestingly, in the in vitro
biochemical assays of TDO-catalyzed NFK formation from L-
tryptophan, an unidentified trace product corresponding to
single oxygen insertion was detected by mass and was proposed
to be 3a-hydroxypyrroloindole-2-carboxylic acid derived from
the unstable epoxide intermediate.14 Thereby, the epoxide is
also most likely to be the intermediate for 2-oxindole formation
via subsequent appropriate epoxide ring opening catalyzed by
MarE.
Due to the lack of an electron-donating group on the indole
ring of 1, the C3−O bond breakage is unlikely to happen via
the pseudo-para-quinone methide formation that takes place in
2-oxindole formation catalyzed by FMOs.4,5 Thus, there are
two plausible ways to open the epoxide ring (3) via the
iminium formation between C2 and the indole nitrogen taking
advantage of the lone electron pair of the indole nitrogen and
the electron deficiency of C2. The first way is summarized in
Scheme 1A. The iminium intermediate (4) is formed via the
C2−O bond breakage and the subsequent attack on 4 by a
water molecule at C2 to yield a 2,3-vicinal diol intermediate
Scheme 1. The Proposed Mechanism of MarE-Catalyzed 2-
Oxindole Formation: (A) The Proposed 2-Oxindole
Formation by Enol−Keto Tautomerization Mechanism
Catalyzed by MarE and (B) The Proposed 2-Oxindole
Formation by 2,3-Hydride Migration Mechanism Used by
MarE
Figure 3. Substrate specificity of MarE. (A) The chemical structures of
putative substrates tested in this study. (B) Kinetic analyses for MarE
with 1, 1b, and 1c were carried out in triplet.
(HR)-MS analysis of the corresponding products confirmed
that a single atom of oxygen was inserted into most of
substrates (Table S2). The products 2-oxindole 2b from 1b and
3-methyl-2-oxindole 2c from 1c were confirmed by comparison
with the authentic standards (Figure S13). Other products were
also most likely 3-substituted 2-oxindoles based on HR-MS
analysis and similar UV−vis spectra to the authentic 2, 2b, and
2c (Figure S14). Among these tested surrogates, MarE showed
higher catalytic activities toward 1, 1b, and 1c. Kinetic analyses
were performed for these three substrates: Km (180.2 17.5
μM) and kcat (0.75 0.02 h−1) for 1, Km (209.5 21.0 μM)
and kcat (2.2 0.07 h−1) for 1b, and Km (303.4 20.4 μM) and
kcat (2.1
0.05 h−1) for 1c (Figure 3B). Although MarE
showed substrate robustness, it could not catalyze the oxidation
of 3′R containing 1h (Figure S13), which is in agreement with
our previous results that showed the incorporation of 1 (2′S,
3′S), not 1h (2′S, 3′R), into maremycines.7 MarE also could
not catalyze the oxidation of the N1-methyl 1i (Figure S13),
which was consistent with the role of MarF as a late-stage N1-
methyltransferase.6c
Mechanical Studies of 2-Oxindole Formation Cata-
lyzed by MarE. TDO-catalyzed oxidation of tryptophan is the
first and rate-limiting step in tryptophan metabolism, playing
crucial roles in regulating the physiological tryptophan levels.
Extensive mechanism studies of TDO-catalyzed oxidation of L-
tryptophan to NFK proposed that the two atoms from
D
DOI: 10.1021/jacs.7b05517
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(5), instead of attacking on 4 by the FeIVO species in the
canonical TDOs (Scheme S1).8 Loss of the hydroxyl group at
C3 via water elimination generates 2-hydroxyindole (6), which
can undergo enol−keto tautomerization to form 2-oxindole 2.
The attack of the iminium by nucleophilic water to generate
2,3-vicinal diol was found in the FqzB assays with tryprostatin A
and B (an early stage precursor of spiro-tryprostatin A and B,
respectively) as the substrates.5
In this route, the hydroxyl group at the C2 originates from
water. To test this hypothesis, the enzymatic reactions were
separately carried out in buffers prepared from H218O and
H216O. However, HR-MS analysis of the product gave the same
mass without the incorporation of 18O into the product (Figure
S15), auguring that the oxygen of C2 may originate from
molecular oxygen. To define the origin of the oxygen atom in 2,
the enzymatic reactions were carried out in a closed chamber at
anaerobic condition or at 18O2-containing condition. No
product was detected for the former reaction (Figure 2E),
while the mass of the product emerged under 18O2-containing
condition was 2 Da greater than that of the unlabeled product
(Figure S16). Therefore, molecular oxygen was demonstrated
to be direct source of 2-oxindoles.
nation-like oxidation of the 3-substituted indoles to generate
the 3-substituted 2-oxindoles, as shown in Scheme 1B.
Biological Basis for the 2-Oxindole formation. Next, we
want to find the different features in the primary sequence of
MarE compared to the typical TDOs. However, although the
multiple sequence alignments of MarE with several well-known
TDOs from Xanthomonas campestris (XcTDO),10 Ralstonia
metallidurans (RmTDO),11 Micromonospora sp. ML1 (TioF),21a
Homo sapiens (HsTDO),21b and Drosophila melanogaster
(DmTDO)21c showed low sequence identity (17.4% with
XcTDO) and moderate similarity (40.9% with XcTDO), they
almost shared the same active sites (Figure S21). The active
sites include H55 (His55 in XcTDO) which is proposed to
form hydrogen-bond interactions with both the indole NH site
of substrate and the predicted FeIVO heme intermedi-
ates,10,22 R118 (Arg117 in XcTDO) which forms bidentate ion-
pair interactions with the carboxyl group of the substrate,10 and
H201 (His240 in XcTDO) that coordinates the iron atom
within the heme cofactor.10,11 Their mutations indeed resulted
in the similar results found for the typical TDOs (Figure
S22).10,11 The circular dichroism (CD) spectra for MarE and
these mutants revealed that they were composed of α-helixes
(typical structural features for TDO superfamily proteins) and
there were no apparent differences in the overall structure
(Figure S23). These observations support MarE as a member of
the TDO superfamily.
Although no apparent differences were found between MarE
and the typical TDOs in the conserved key amino acid residues,
a phylogenetic tree was constructed to include MarE and its
homologues that were extracted from the NCBI by blastp.
MarE was found to locate in a separate clade from the other
typical TDOs (Figure S24). ACPL_6188 (45.8% identity to
MarE) from Actinoplanes sp. SE50/11023 locates in the same
clade with MarE. To test whether ACPL_6188 has the same
catalytic activity, this protein was cloned, expressed, and
purified to homogeneity from E. coli BL21 (DE3). When
ACPL_6188 was incubated with ascorbate and 1, 1b, 1c, 1d,
1e, 1f, and 1g respectively, the corresponding 2-oxindole
products (2, 2b, 2c, 2d, and 2e) were produced albeit in lower
catalytic efficiency in comparison with MarE, but no apparent
products were formed from 1f or 1g (Figure S25). Thus, MarE,
ACPL_6188, and other homologues may undergo a specific
evolutionary process to obtain this new catalytic function,
specific for natural product biosynthesis. To clarify why MarE
catalyzes the formation of 2-oxindole by insertion of a single
oxygen atom instead of an oxidative cleavage by insertion of
two oxygen atoms, more structural information about protein
binding the substrate or its analogs or inhibitors may be
needed. This study is underway.
The labeling studies suggest another mechanism in which
hydride migration from the C2 to C3 of the epoxide (3) can
directly lead to the 2-oxindole (Scheme 1B). In this process, the
transient iminium intermediate (8) is generated by concom-
itant hydride transfer from C2 to C3 and epoxide opening via
the C3−O bond breakage, so the hydrogen at C3 of the
product 2 originating from C2 will be retained.19 To test this
hypothesis, the enzymatic reactions were carried out with 1 as
the substrate in D2O buffer at pH 7.0, not pH 8.0 previously
used for assays in order to slow down the keto−enol
tautomerization of 2. Liquid chromatography (LC)-MS was
used to monitor whether the hydrogen at C3 of 2 is retained or
lost. At three different time points (0.5, 2.0, and 12.0 h), both 2
and deuterated 2 were detected, but the ratio of 2 to the
deuterated 2 decreased along with the prolonged reaction time
(Figure S17). 3-Methylindole (1c) was also used as a substrate
instead of 1 in the reactions of MarE in D2O buffer. In the three
time points (0.5, 1.0, and 4.0 h), both 2c and deuterated 2c
were detected, but 2c was the major product in the 0.5 h
reaction and the ratio of 2c to deuterated 2c also decreased
along the incubation time (Figure S18). These results indicated
that the hydrogen at C3 indeed originates from C2 and the
keto−enol tautomerization of 2 still took place relatively quick,
compared to slower MarE-catalyzed 2-oxindole formation.
To further confirm this result, [2,4,5,6,7-D5]-(2′S,3′S)-
methyltryptophan (1a) was synthesized according to a
literature procedure.20 The degree of deuteration was measured
1
by H NMR spectroscopy and ranged from 78% at C4 to 99%
DISCUSSION
at C2 and C5 (Table S3 and Figure S19). With 1a (Figure 3A)
as the substrate, the enzymatic assays were carried out in H2O-
buffer at pH 7.0. LC-MS analysis showed that 2a (the D5-
labeled product) was the dominant product in the 20 min
reaction, and the content of 2a in the reaction mixture similarly
decreased for longer reaction time (Figure S20). These results
solidly verified that the 2-oxindole structure was first formed in
the catalysis of MarE via hydride migration from the C2 to C3
of the epoxide and then underwent the spontaneous keto−enol
tautomerization in buffer, resulting in the exchange of hydrogen
and deuterium at C3. Thus, a conclusion could be made that
the formation of an epoxide and subsequent 2,3-hydride
migration were utilized by MarE to catalyze the monooxyge-
■
Although playing different roles in the physiological conditions
and forming different natural polymer states, TDOs (homote-
tramer) and IDOs (monomer) both utilize the FeII/heme
cofactor to activate O2 for the formation of NFK by
dioxygenatively splitting the indole ring of tryptophan. The
catalytic mechanisms of TDOs and IDOs are still not absolutely
clear, mainly because no direct intermediates were obtained or
solidly characterized.8 In early studies, Criegee and dioxetane
mechanisms were put forward, and both of them proposed that
the two atoms of O2 were inserted into the indole ring of
tryptophan at the same time.14 However, these two
mechanisms were ruled out by later crystallographic,10,11,21
E
DOI: 10.1021/jacs.7b05517
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
computational,17,18 and spectroscopic14 studies. Subsequently, a
new mechanism was proposed that the two atoms of O2 are
sequentially inserted into the indole ring upon first forming a
key 2,3-epoxide intermediate. However, the proposed epoxide
intermediate has not been directly isolated due to the lability of
the epoxy ring.
In this study, we characterized MarE as a 2-oxindole motif
forming TDO homologue in the biosynthetic pathway of
maremycins by in vivo genetic studies and in vitro biochemical
assays. MarE contains most of the key amino acid residues
conserved among the TDO superfamily members and forms
the similar homotetramer in the solution. However, MarE
catalyzes the insertion of a single atom of O2 to produce a 2-
oxindole structure like a monooxygenase, not the canonical
insertion of two oxygen atoms for the cleaved product NFK
formation.
ization of MarE expanded the functional diversity of TDO
superfamily.
In the canonical TDO-catalyzed reaction cycle (Scheme S1),
heme-FeIII should be first reduced into its ferrous (FeII) state
for O2 activation because O2 is not a reactive molecule from the
point view of chemistry,26 and then FeII−O2 or FeIII−O−O•−
attacks the double bond of the indole to form the 2,3-
epoxyindole together with the ferryl species (FeIVO) formed,
and finally the oxidative FeIVO attacks the 4-like iminium
intermediate, resulting in NFK and FeIII 8. During this process,
reducing agents, such as ascorbate, play roles in reducing the
ferric (FeIII) state of the enzyme into its ferrous (FeII) state to
start the catalytic cycle. In this study, no product 2 was formed
from 1 without reducing agents (Figure 2E), so MarE should
also utilize FeII to activate O2 molecules in the formation of 2-
oxindoles. Distinctly, the formed ferryl species (FeIVO)
cannot attack the iminium intermediate 8 probably due to the
presence of the oxygen anion, but will be directly reduced to
FeII and H2O by the reducing agents (ascorbate or GSH in this
study), with the newly formed ferrous (FeII) state enzyme
entering another catalytic cycle (Scheme 1B).
In the previous Criegee and dioxetane proposals, an
enzymatic base was needed for deprotonation of the indole
NH group in the substrate, and a conserved histidine (H55 in
XcTDO) was assigned for this function.8 However, the later
studies found that 1-methyl-L-tryptophan was still a slow
substrate for XcTDO-H55A and XcTDO-H55S mutants.27
Although the hydrogen-bond interaction was found between
H55 and NH group of L-tryptophan in the structure of
XcTDO,10 the proposed deprotonation step was confirmed to
be not needed. In addition, HsIDO does not have a conserved
histidine at all.28 MarE has the same conserved H55 as XcTDO
does, the mutation of which significantly lost the activity. No
product was detected when 1-methyl-L-tryptophan (1i) was
used as a putative substrate for MarE (Figure S13). These
results seem to suggest that H55 acts as a deprotonation base.
But given that the conversion of L-tryptophan (1d) is very slow,
the reaction of 1i may be too slow to detect its product.
Thereby, 1-methylindole 1j was also used to examine the
function of H55. Although it was also a slow substrate
compared with indole (1b), the product 2j could be obviously
detected by HPLC (Figure S13) and confirmed by HR-MS
(Table S2). Based on all these results, H55 in MarE can be
determined to play the roles in forming hydrogen-bond
interactions both with the NH group of 1 and heme-bound
oxygen like the typical TDOs,10,22,29 not as the deprotonation
base. MarE cannot catalyze the oxygenation of 1i, most
probably because of the spatial hindrance of MarE. This result
is also consistent with the biosynthetic pathway of maremycins
that N1-methylation of the indole moiety by MarF is later than
MarE-catalyzed oxygenation.6c
Previously, IDO has been found to catalyze both
monooxygenation and dioxygenation of indole with H2O2 as
the oxidant, generating multiple products including 2-oxindole
and 3-oxindole.24 H2O2 was determined as the sole source of
the oxygen of products, but the mechanism for the formation of
2-oxindole and 3-oxindole remains unclear. Based on our
current data, IDO may also employ an epoxide intermediate for
2-oxindole and 3-oxindole formation. However, the epoxide
ring opening by the addition of water (Scheme 1A) or hydride
migration (Scheme 1B) remains obscure. Since both 2-oxindole
and 3-oxindole were formed, it would be more reasonable that
the epoxide ring is opened by water and subsequently
undergoes the dehydration and enol−keto tautomerization, as
proposed by us (Scheme 1A). Compared to the IDO-catalyzed
indole oxygenation with H2O2, MarE is very unique that it
specifically catalyzes the oxygenation of both indole and a series
of 3-substituted indole analogs to produce their corresponding
2-oxindoles with O2 as the oxidant.
PrnB is also an IDO homologue from the pyrrolnitrin
biosynthetic pathway and has been genetically and structurally
proposed to catalyze the monooxygenation-like reaction in the
complicated transformation of 7-chloro tryptophan into
monodechloroamino-pyrrolnitrin.25 The structural studies
proposed that the formed iron peroxy species favors to attack
the C3 position to perform the hydroxylation at C3, and the
iminium intermediate is formed with C2, because the oxygen is
very close to the C3.25c,d In the case of MarE, the formed iron
peroxy species may also first attack the C3, consistent with the
indole chemistry, but the epoxide is subsequently formed,
rather than the hydroxylation at the C3. If the same
intermediate is formed in the MarE-catalyzed reaction, the 2-
oxindole must be produced by the dehydration and enol−keto
tautomerization mechanism (Scheme 1A). Recently, 3a-
hydroxy pyrroloindole-2-carboxylic acid detected in some
biochemical assays of TDOs and IDOs is proposed to be
formed by breaking the C2−O bond of the epoxide to generate
the 4-like iminium intermediate that undergoes the nucleophilic
attack of the amino group.14 Along with our data and others,14
these results indicate that the epoxide is also a possible
intermediate in the PrnB-catalyzed reactions. Thus, the
epoxyindole ring opening by cleavage of the C2−O bond is
the most common pattern in heme-dependent dioxygenase,8
while it is unprecedented that the ring is opened by 2,3-hydride
migration resulting in the C3−O bond breakage in the MarE-
catalyzed reaction. MarE belongs to the TDO family (no
homology to both PrnB and HsIDO), whereas PrnB belongs to
IDO family. Therefore, the catalytic mechanism character-
In conclusion, MarE and ACPL_6188, members of
tryptophan 2,3-dioxygenase superfamily, have been biochemi-
cally characterized to catalyze the unique mono-oxygenation of
the substrate under reducing conditions, not canonical
dioxygenative indole ring cleavage. They showed excellent
substrate promiscuity, catalyzing the formation of a series of 3-
substituted 2-oxindoles, the common motifs found in many
natural products. This study expands the catalytic diversity of
TDO superfamily. MarE and its homologues could serve as new
genome mining probes for the discovery of novel bioactive
natural products containing the 2-oxindole pharmacophore.
F
DOI: 10.1021/jacs.7b05517
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
((v/v): 80:20 to 0:100, 0−18 min; 0:100, 18−20 min) with
detection at 254 nm.
METHODS AND EXPERIMENTAL SECTION
■
Materials, Strains, and Plasmids. Primers were synthe-
sized by Shanghai Sangon Biotech Co. Ltd. (China). Standard
chemical reagents were purchased from Sigma-Aldrich
Corporation (USA), Aladdin Industrial Corporation (China),
and Merck KGaA (Germany). The wild-type Streptomyces sp.
B9173, ΔmarE mutant (LS25), and marE gene complement
mutant strain (ZYY1) were grown in solid mannitol soya flour
medium (MS medium, mannitol 2%, soya flour 2%, agar 2%) at
30 °C for spore production and conjugation and ISP2 medium
(malt extract 1%, yeast extract 0.4%, glucose 0.4%) at 30 °C for
maremycins fermentation. The E. coli strains were grown in
Luria−Bertani medium (LB medium, tryptone 1%, yeast extract
0.5%, and sodium chloride 0.5%) for gene manipulation and
protein expression. Kanamycin (50 μg/mL), chloramphenicol
(50 μg/mL), and apramycin (50 μg/mL) were used for
selection of recombinant strains. The primers, plasmids, and
strains used in this study were summarized in Tables S4−S6.
In Vitro Biochemical Assays and Product Analysis. The
reaction mixture (100 μL) for MarE or ACPL_6188 contained
10 μM enzyme, 10 mM L-ascorbate, and 1 mM substrate, such
as (2′S,3′S)-methytryptophan (1), indole (1b), 3-methylindole
(1c), L-tryptophan (1d), D-tryptophan (1e), tryptamine (1f),
indole-3-propionic acid (1g), (2′S, 3′R)-methyltryptophan
(1h), 1-methyl-L-tryptophan (1i), or 1-methyl-indole (1j), in
50 mM HEPES buffer (pH 8.0) and was incubated at 28 °C for
6 h. Reactions were quenched by being frozen in −80 °C
quickly and then lyophilizing. 100 μL methanol was used to
redissolve the reaction mixture, and 10 μL supernatant after
centrifugation (12000 rpm, 10 min) to remove the proteins was
subjected to HPLC analysis.
The column used for HR-LC-Q-TOF analysis was Agilent
Eclipse XDB-C18 column (5 μm, 4.6 × 150 mm). The flow
rate was 0.3 mL/min. The mobile phase was comprised of
water and acetonitrile containing 0.1% (v/v) formic acid ((v/
v): 90:10 to 60:40, 0−15 min; 60:40 to 0:100, 15−20 min;
0:100, 20−25 min) with detection at 254 nm.
HPLC was performed on an Agilent HPLC Series 1200. LC-
ESI-MS was performed on Agilent 1100 series LC/MSD Trap
in positive mode. HR-LC-MS was performed on Agilent 1200
series LC-MS system with a 6530 Q-TOF mass spectrometer in
positive mode.
Preparative Scale in Vitro Reaction for Structural
Elucidation of 2. A solution with a total volume of 100 mL
composed of 50 μM MarE, 1 mM 1, and 10 mM L-ascorbate in
the HEPES buffer (50 mM, pH 8.0) was incubated at 28 °C.
The reaction was monitored by HPLC analysis. After 24 h
incubation, the reaction was quenched by being frozen at −80
°C and then lyophilized to dryness. One ml 10% (v/v)
methanol was used to redissolve the reaction mixture, and the
supernatant after centrifugation (15000 rpm, 10 min) was used
to purify the products. The column used for purification was
flash column packed with reverse phase silica gel (YMG*GEL,
ODS-A-HG, 50 μm, 2 × 15 cm). The purity of 2 was checked
by HPLC analysis.
1
1
For structural elucidation, H, 13C, H−1H COSY, HMBC,
and HSQC spectra were recorded on Bruker AV 600
instruments.
Chemical Synthesis of 1a. 1a was directly obtained from
(2′S,3′S)-methyltryptophan 1 according to the reported
method20 with several modifications. In one capped glass
flask, 10 mg 1 was dissolved in 600 μL D2O/CF3COOD (1:2,
v/v) mixture. The flask was wrapped up with aluminum foil and
stirred with a magnetic stirrer for 3 days at room temperature.
When the reactions were conducted in H218O containing
buffer, MarE was stored in the corresponding buffer after
purification in the normal buffer. The reactions were incubated
at 28 °C for 6 h and quenched by adding 100 μL methanol to
precipitate the proteins. The supernatants after centrifugation
(12000 rpm, 10 min) were subjected to HR-LC-Q-TOF
analysis.
1
1a was dried and checked by LC-HR-ESI-MS and H NMR.
The sample was deuterated for 7 times as the procedure above
until most of the hydrogen atoms on the indole ring of 1 were
exchanged with D atoms, especially the C2 of indole ring
reaching about 100% deuteration.
When the reactions were conducted under anaerobic
condition or 18O2-containing condition, a closed chamber
with the previously degassed reaction system (20 μM MarE, 2
mM 1, 10 mM ascorbate in 50 mM HEPES buffer (pH 8.0))
was rapidly evacuated first and then filled with pure N2. This
process was checked by a rubber balloon and successively done
three times. Then the evacuated chamber was filled with 18O2,
and the reactions were performed for 6 h at 28 °C. The
reaction system as negative control was not filled with 18O2.
When the reactions were conducted at D2O-made buffer (pH
7.0) for 1 and 1c or H2O-made buffer (pH 7.0) for 1a, MarE
was purified and stored at the corresponding buffers. The
reactions were incubated at 28 °C for different time points, and
the supernatants were subjected to HPLC-ESI-MS analysis after
centrifugation (15000 rpm, 1 min).
The Experimental Procedures and Methods. Details for
gene inactivation, complementation, and metabolites analysis;
cloning, expression, purification, analysis of recombinant
proteins, and kinetic analysis; and bioinformatic analysis and
phylogenetic analysis are described in the Supporting
Information.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.7b05517.
Experimental procedures, HPLC data, HR-MS data,
NMR data, sequence analysis, including Figures S1−S25,
Tables S1−S6 and Scheme S1 (PDF)
The column used for HPLC and HPLC-ESI-MS detection
was a C18 column (10 μm, 4.6 × 200 mm). The flow rate was
0.5 mL/min for HPLC and 0.4 mL/min for LC-ESI-MS. The
mobile phase for 1, 1d, 1e, 1f, 1g, 1h, and 1i was comprised of
water and acetonitrile containing 0.05% (v/v) formic acid ((v/
v): 95:5 to 60:40, 0−15 min; 60:40 to 0:100, 15−20 min;
0:100, 20−25 min) with detection at 254 nm. The mobile
phase for 1b, 1c, and 1j was comprised of water and acetonitrile
AUTHOR INFORMATION
■
Corresponding Author
*linsj@sjtu.edu.cn
ORCID
Yi Tang: 0000-0003-1597-0141
Shuangjun Lin: 0000-0001-9406-9233
G
DOI: 10.1021/jacs.7b05517
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
D.; Parmentier, N.; Boon, T.; Van den Eynde, B. J. Nat. Med. 2003, 9,
1269.
Notes
The authors declare no competing financial interest.
(10) Forouhar, F.; Anderson, J. L.; Mowat, C. G.; Vorobiev, S. M.;
Hussain, A.; Abashidze, M.; Bruckmann, C.; Thackray, S. J.;
Seetharaman, J.; Tucker, T.; Xiao, R.; Ma, L. C.; Zhao, L.; Acton, T.
B.; Montelione, G. T.; Chapman, S. K.; Tong, L. Proc. Natl. Acad. Sci.
U. S. A. 2007, 104, 473.
(11) Zhang, Y.; Kang, S. A.; Mukherjee, T.; Bale, S.; Crane, B. R.;
Begley, T. P.; Ealick, S. E. Biochemistry 2007, 46, 145.
(12) Fu, R.; Gupta, R.; Geng, J.; Dornevil, K.; Wang, S.; Zhang, Y.;
Hendrich, M. P.; Liu, A. J. Biol. Chem. 2011, 286, 26541.
(13) (a) Batabyal, D.; Yeh, S. R. J. Am. Chem. Soc. 2007, 129, 15690.
(b) Li, J. S.; Han, Q.; Fang, J.; Rizzi, M.; James, A. A.; Li, J. J. Arch.
Insect Biochem. Physiol. 2007, 64, 74−87. (c) Basran, J.; Rafice, S. A.;
Chauhan, N.; Efimov, I.; Cheesman, M. R.; Ghamsari, L.; Raven, E. L.
Biochemistry 2008, 47, 4752.
ACKNOWLEDGMENTS
■
We thank Mrs. B. Dai and J. Wu at the Instrumental Analysis
Center of SJTU for making NMR experiments. We also thank
Prof. H.-W. Liu at University of Texas at Austin for helpful
discussion and proofreading, and Prof. L. Bai at SJTU for
providing Actinoplanes sp. SE50/110 strain. We thank HKI,
Jena (Germany) and Prof. Osamu Tamura from Showa
Pharmaceutical University, Tokyo (Japan) for providing the
Streptomyces sp. B9173 strain and maremycin A standard. This
work was financially supported by the National Science
Foundation of China (31425001, 21632007, and
21661140002) and National Key Research and Development
Program of China (grant no. 2016YFA0601104). This study is
a contribution to the international IMBER project.
(14) Basran, J.; Efimov, I.; Chauhan, N.; Thackray, S. J.; Krupa, J. L.;
Eaton, G.; Griffith, G. A.; Mowat, C. G.; Handa, S.; Raven, E. L. J. Am.
Chem. Soc. 2011, 133, 16251.
(15) Forman, H. J.; Zhang, H.; Rinna, A. Mol. Aspects Med. 2009, 30,
1.
(16) We thank one of the reviewers for this experiment and the
subsequent insight towards interpreting the results. Both glutathione
and dithionite inhibited the reaction and formed a potential hexa-
coordinating low-spin heme species.
(17) Lewis-Ballester, A.; Batabyal, D.; Egawa, T.; Lu, C.; Lin, Y.;
Marti, M. A.; Capece, L.; Estrin, D. A.; Yeh, S. R. Proc. Natl. Acad. Sci.
U. S. A. 2009, 106, 17371.
(18) Chung, L. W.; Li, X.; Sugimoto, H.; Shiro, Y.; Morokuma, K. J.
Am. Chem. Soc. 2010, 132, 11993.
(19) Skordos, K. W.; Skiles, G. L.; Laycock, J. D.; Lanza, D. L.; Yost,
G. S. Chem. Res. Toxicol. 1998, 11, 741.
REFERENCES
■
(1) (a) Peddibhotla, S. Curr. Bioact. Compd. 2009, 5, 20. (b) Yan, T.;
Wang, X.; Sun, H.; Liu, J.; Xie, Y. Molecules 2013, 18, 14505. (c) Singh,
S. G.; Desta, Y. Z. Chem. Rev. 2012, 112, 6104. (d) Yu, B.; Yu, D. Q.;
Liu, H. M. Eur. J. Med. Chem. 2015, 97, 673. (e) Kaur, M.; Singh, M.;
Chadha, N.; Silakari, O. Eur. J. Med. Chem. 2016, 123, 858. (f) Ball-
Jones, N. R.; Badillo, J. J.; Franz, A. K. Org. Biomol. Chem. 2012, 10,
5165. (g) Antonchick, A. P.; Gerding-Reimers, C.; Catarinella, M.;
Schurmann, M.; Preut, H.; Ziegler, S.; Rauh, D.; Waldmann, H. Nat.
Chem. 2010, 2, 735.
(2) (a) Rottmann, M.; McNamara, C.; Yeung, B. K.; Lee, M. C.; Zou,
B.; Russell, B.; Seitz, P.; Plouffe, D. M.; Dharia, N. V.; Tan, J.; Cohen,
S. B.; Spencer, K. R.; Gonzalez-Paez, G. E.; Lakshminarayana, S. B.;
́ ́
(20) Winnicka, E.; Kan
675.
́
ska, M. J. Radioanal. Nucl. Chem. 2009, 279,
(21) (a) Sheoran, A.; King, A.; Velasco, A.; Pero, J. M.; Garneau-
Tsodikova, S. Mol. BioSyst. 2008, 4, 622. (b) Meng, B.; Wu, D.; Gu, J.;
Ouyang, S.; Ding, W.; Liu, Z. J. Proteins: Struct., Funct., Genet. 2014, 82,
3210. (c) Huang, W.; Gong, Z.; Li, J.; Ding, J. J. Struct. Biol. 2013, 181,
291.
(22) Geng, J.; Dornevil, K.; Liu, A. J. Am. Chem. Soc. 2012, 134,
12209.
(23) Schwientek, P.; Szczepanowski, R.; Ruckert, C.; Kalinowski, J.;
Klein, A.; Selber, K.; Wehmeier, U. F.; Stoye, J.; Puhler, A. BMC
Genomics 2012, 13, 112.
Goh, A.; Suwanarusk, R.; Jegla, T.; Schmitt, E. K.; Beck, H. P.; Brun,
R.; Nosten, F.; Renia, L.; Dartois, V.; Keller, T. H.; Fidock, D. A.;
Winzeler, E. A.; Diagana, T. T. Science 2010, 329, 1175. (b) Tomita,
D.; Yamatsugu, K.; Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2009,
131, 6946. (c) Bill, K. L.; Garnett, J.; Meaux, I.; Ma, X.; Creighton, C.
J.; Bolshakov, S.; Barriere, C.; Debussche, L.; Lazar, A. J.; Prudner, B.
C.; Casadei, L.; Braggio, D.; Lopez, G.; Zewdu, A.; Bid, H.; Lev, D.;
Pollock, R. E. Clin. Cancer Res. 2016, 22, 1150.
(3) (a) Goodman, V. L.; Rock, E. P.; Dagher, R.; Ramchandani, R. P.;
Abraham, R. P. S.; Gobburu, J. V.; Booth, B. P.; Verbois, S. L.; Morse,
D. E.; Liang, C. Y.; et al. Clin. Cancer Res. 2007, 13, 1367. (b) London,
C. A.; Malps, P. B.; Wood-Follis, S. L.; Boucher, J. F.; Rusk, A. W.;
Rosenberg, M. P.; Henry, C. J.; Mitchener, K. L.; Klein, M. K.;
Hintermeister, J. G.; et al. Clin. Cancer Res. 2009, 15, 3856.
(4) Li, S.; Finefield, J. M.; Sunderhaus, J. D.; McAfoos, T. J.;
Williams, R. M.; Sherman, D. H. J. Am. Chem. Soc. 2012, 134, 788.
(5) Tsunematsu, Y.; Ishikawa, N.; Wakana, D.; Goda, Y.; Noguchi,
H.; Moriya, H.; Hotta, K.; Watanabe, K. Nat. Chem. Biol. 2013, 9, 818.
(6) (a) Bindseil, B. W.; Helmke, E.; Weyland, H.; Laatsch, H. Liebigs
Ann. Chem. 1995, 1995, 1291. (b) Tang, Y.; Sattler, I.; Thiericke, R.;
Grabley, S.; Feng, X. Eur. J. Org. Chem. 2001, 2001, 261. (c) Lan, Y.;
Zou, Y.; Huang, T.; Wang, X.; Brock, L. N.; Deng, Z.; Lin, S. Sci.
China: Chem. 2016, 59, 1224.
(24) Kuo, H. H.; Mauk, A. G. Proc. Natl. Acad. Sci. U. S. A. 2012, 109,
13966.
(25) (a) Kirner, S.; Hammer, P. E.; Hill, D. S.; Altmann, A.; Fischer,
I.; Weislo, L. J.; Lanahan, M.; van Pee, K. H.; Ligon, J. M. J. Bacteriol.
1998, 180, 1939. (b) De Laurentis, W.; Leang, K.; Hahn, K.;
Podemski, B.; Adam, A.; Kroschwald, S.; Carter, L. G.; van Pee, K. H.;
Naismith, J. H. Acta Crystallogr., Sect. F: Struct. Biol. Cryst. Commun.
2006, 62, 1134. (c) De Laurentis, W.; Khim, L.; Anderson, J. L.; Adam,
A.; Johnson, K. A.; Phillips, R. S.; Chapman, S. K.; van Pee, K. H.;
Naismith, J. H. Biochemistry 2007, 46, 12393. (d) Zhu, X.; van Pee, K.
H.; Naismith, J. H. J. Biol. Chem. 2010, 285, 21126.
(26) Poulos, T. L. Chem. Rev. 2014, 114, 3919.
(27) Chauhan, N.; Thackray, S. J.; Rafice, S. A.; Eaton, G.; Lee, M.;
Efimov, I.; Basran, J.; Jenkins, P. R.; Mowat, C. G.; Chapman, S. K.;
Raven, E. L. J. Am. Chem. Soc. 2009, 131, 4186.
(28) Chauhan, N.; Basran, J.; Efimov, I.; Svistunenko, D. A.; Seward,
H. E.; Moody, P. C.; Raven, E. L. Biochemistry 2008, 47, 4761.
(29) Davydov, R. M.; Chauhan, N.; Thackray, S. J.; Anderson, J. L.;
Papadopoulou, N. D.; Mowat, C. G.; Chapman, S. K.; Raven, E. L.;
Hoffman, B. M. J. Am. Chem. Soc. 2010, 132, 5494.
(7) Zou, Y.; Fang, Q.; Yin, H.; Liang, Z.; Kong, D.; Bai, L.; Deng, Z.;
Lin, S. Angew. Chem., Int. Ed. 2013, 52, 12951.
(8) (a) Efimov, I.; Basran, J.; Thackray, S. J.; Handa, S.; Mowat, C.
G.; Raven, E. L. Biochemistry 2011, 50, 2717. (b) Geng, J.; Liu, A. Arch.
Biochem. Biophys. 2014, 544, 18.
(9) (a) Pilotte, L.; Larrieu, P.; Stroobant, V.; Colau, D.; Dolusic, E.;
Frederick, R.; De Plaen, E.; Uyttenhove, C.; Wouters, J.; Masereel, B.;
Van denEynde, B. J. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 2497.
(b) Breda, C.; Sathyasaikumar, K. V.; SograteIdrissi, S.; Notarangelo,
F. M.; Estranero, J. G.; Moore, G. G.; Green, E. W.; Kyriacou, C. P.;
Schwarcz, R.; Giorgini, F. Proc. Natl. Acad. Sci. U. S. A. 2016, 113,
5435. (c) Uyttenhove, C.; Pilotte, L.; Theate, I.; Stroobant, V.; Colau,
H
DOI: 10.1021/jacs.7b05517
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX