Molecular and Structural Characterization of a Promiscuous C-Glycosyltransferase from Trollius chinensis

Article abstract of DOI:10.1002/anie.201905505

Herein, the catalytic promiscuity of TcCGT1, a new C-glycosyltransferase (CGT) from the medicinal plant Trollius chinensis is explored. TcCGT1 could efficiently and regio-specifically catalyze the 8-C-glycosylation of 36 flavones and other flavonoids and could also catalyze the O-glycosylation of diverse phenolics. The crystal structure of TcCGT1 in complex with uridine diphosphate was determined at 1.85 ? resolution. Molecular docking revealed a new model for the catalytic mechanism of TcCGT1, which is initiated by the spontaneous deprotonation of the substrate. The spacious binding pocket explains the substrate promiscuity, and the binding pose of the substrate determines C- or O-glycosylation activity. Site-directed mutagenesis at two residues (I94E and G284K) switched C- to O-glycosylation. TcCGT1 is the first plant CGT with a crystal structure and the first flavone 8-C-glycosyltransferase described. This provides a basis for designing efficient glycosylation biocatalysts.

Full text of DOI:10.1002/anie.201905505

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Molecular Characterization and Structural Basis of a
Promiscuous C-Glycosyltransferase from Trollius chinensis
Authors: Jun-Bin He, Peng Zhao, Zhi-Min Hu, Shuang Liu, Yi Kuang,
Meng Zhang, Bin Li, Cai-Hong Yun, Xue Qiao, and Min Ye
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: Angew. Chem. Int. Ed. 10.1002/anie.201905505
Angew. Chem. 10.1002/ange.201905505
Link to VoR: http://dx.doi.org/10.1002/anie.201905505
http://dx.doi.org/10.1002/ange.201905505

10.1002/anie.201905505
Angewandte Chemie International Edition
RESEARCH ARTICLE
Molecular Characterization and Structural Basis of a Promiscuous
C-Glycosyltransferase from Trollius chinensis
Jun-Bin He,†,[a] Peng Zhao,†,[b] Zhi-Min Hu,[a] Shuang Liu,[a] Yi Kuang,[a] Meng Zhang,[a] Bin Li,[a] Cai-Hong
Yun,*[b] Xue Qiao,*[a] and Min Ye*[a]
Abstract: In this work, we explored the catalytic promiscuity of
TcCGT1, a new C-glycosyltransferase (CGT) from the medicinal plant
Trollius chinensis. TcCGT1 could efficiently and regio-specifically
catalyze 8-C-glycosylation of 36 flavones and other flavonoids, and
could also catalyze the O-glycosylation of diverse phenolics.
Moreover, the crystal structure of TcCGT1 in complex with uridine
diphosphate was determined at 1.85 Å resolution. Structural analysis
with molecular docking revealed a new model for catalytic mechanism
of TcCGT1, which was initiated by substrate spontaneous
deprotonation. The spacious binding pocket explains the robust
substrate promiscuity, and binding pose of the substrate determines
C- or O-glycosylation activity. Site-directed mutagenesis at two
residues (I94E and G284K) switched C- to O-glycosylation. This work
highlights TcCGT1 as the first plant CGT with a crystal structure and
the first flavone 8-C-glycosyltransferase, and provides a basis for
protein engineering to design efficient glycosylation biocatalysts for
drug discovery.
glycosylation reactions mediated by enzymes, i.e. C-
glycosyltransferases (CGTs), show high efficiency and region-
specificity.[7]
Recently, the enzymatic synthesis of flavonoid C-glycosides by
plant CGTs has attracted considerable interest (Scheme S1).[7,8]
The 2-hydroxyflavanone CGTs, including OsCGT from Oryza
sativa,[8a] FeCGTa and FeCGTb from Fagopyrum esculentum,[8d]
UGT708D1 from Glycine max,[8f] FcCGT from Fortunella
crassifolia, and CuCGT from Citrus unshiu,[8i] utilize the open-
circular form of 2-hydroxyflavanones as substrates to produce
flavone C-glycosides after dehydration. The bifunctional
glycosyltransferase (GT) UGT708A6 from Zea mays uses the
close-circular form of 2-hydroxyflavanones as substrates.[8c]
Moreover, GtUF6CGT1 from Gentiana triflora could directly
catalyze the 6-C-glycosylation of flavones, and PlUGT43 from
Pueraria
lobata
catalyzes
the
8-C-glycosylation
of
isoflavones.[8e,8h] However, no CGTs have been discovered, so far,
to catalyze the 8-C-glycosylation of flavones to produce important
bioactive natural products like 1a and 2a. Moreover, the known
CGTs exhibit relatively low substrate promiscuity and poor
catalytic efficiency, which limit their application in the synthesis of
structurally diverse C-glycosides. MiCGT recently reported from
Mangifera indica could accept a variety of natural and unnatural
Introduction
Glycosides are a big class of bioactive natural products.[1]
Compared with other types of glycosides (O-, N-, or S-glycosides),
C-glycosides are stable to enzymatic or chemical hydrolysis due
to the rigid C–C bond between the sugar residue and the
aglycone.[1,2] Flavonoid C-glycosides, which are widely present in
plants, exhibit significant benefits to human health.[3] Typical
examples include vitexin (apigenin 8-C-β-D-glucoside, 1a) and
orientin (luteolin 8-C-β-D-glucoside, 2a), which show anti-oxidant,
anti-cancer, anti-inflammatory, antiviral, antimicrobial, and anti-
diabetic activities.[3,4] Different strategies to synthesize flavonoid
C-glycosides have been explored.[5] Chemical synthesis is usually
restricted by low yield, poor selectivity, and multi-steps of
protection and deprotection of functional groups.[6] In contrast, C-
substrates, but majority of them feature
a
2,4,6-
trihydroxybenzophenone-like core structure.[7c,9] Thus, it is critical
to mine novel CGTs with catalytic promiscuity to improve
structural diversity of natural products for drug discovery.
The catalytic mechanisms of O-glycosyltransferases (OGTs)
and bacterial CGTs have been extensively studied on the basis of
crystal structures.[10] For instance, a highly conserved catalytic
dyad (His-Asp) arrangement of active-site residues are essential
for glycosylation activities of plant OGTs.[10a-f] In the case of
bacterial CGTs, Asp137 in UrdGT2, or Asp58 and Glu316 in
SsfS6 serves as a catalytic base to accept an aromatic proton of
the acceptors, thereby facilitating the nucleophilic attack at the
sugar anomeric carbon to form a C–C bond.[10g,10h] However, no
crystal structures have been reported for plant CGTs, so far. A
few groups have tried to elucidate their catalytic mechanisms
based on protein modeling and site-directed mutagenesis. Hirade
et al. reported that Asp85 and Arg292 located in the active site of
UGT708D1 were critical for the C-glucosylation activity.[8f] Chen
et al. proposed that Ile152 was the critical amino acid residue for
di-C-glycosylation of MiCGTb.[11] Gutmann and Nidetzky found
the exchange of active-site motif (Ile-Asp or Asp-Ile) in PcOGT
and OsCGT could achieve interconversion of O- and C-
glycosylation.[12] Nevertheless, crystal structure information would
remarkably improve understanding on the glycosylation
mechanisms of plant CGTs.
[a]
J. -B. He, Z. -M. Hu, S. Liu, Y. Kuang, M. Zhang, B. Li, Dr. X. Qiao,*
Prof. M. Ye*
State Key Laboratory of Natural and Biomimetic Drugs & Key
Laboratory of Molecular Cardiovascular Sciences of Ministry of
Education, School of Pharmaceutical Sciences, Peking University,
38 Xueyuan Road, Beijing 100191, China
E-mail: qiaoxue@bjmu.edu.cn; yemin@bjmu.edu.cn
Dr. P. Zhao, Prof. C. -H. Yun*
Department of Biochemistry and Biophysics & Department of
Integration of Chinese and Western Medicine, School of Basic
Medical Sciences, Peking University, 38 Xueyuan Road, Beijing
100191, China
[b]
[†]
E-mail: yunch@hsc.pku.edu.cn
These authors contributed equally to this work.
Trollius chinensis Bunge (Ranunculaceae) is an endemic plant
in China. Its flowers are widely used in traditional Chinese
Supporting information, including detailed description of the
experimental methods and additional experimental data, 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/anie.201905505
Angewandte Chemie International Edition
RESEARCH ARTICLE
C-glycosides.[16] The structure of 1a was fully identified as vitexin
by comparing with an authentic reference standard. Likewise, 2a
was identified as orientin. These results unequivocally
established TcCGT1 as a CGT that catalyzes the C-glycosylation
at C-8 of flavones. To our best knowledge, TcCGT1 is the first
identified flavone 8-C-glycosyltransferase.
medicine due to significant antiviral, antibacterial, anti-
inflammatory, antioxidant, and anticancer activities.[13] Flavone C-
glycosides including vitexin (1a) and orientin (2a) are its major
bioactive compounds,[14] indicating the existence of CGTs in this
plant. Nevertheless, no CGTs have been reported to synthesize
flavone 8-C-glycosides.[15] Herein, we report
a novel C-
glycosyltransferase TcCGT1 from T. chinensis, which represents
the first CGT to catalyze 8-C-glycosylation of flavones. We also
elucidated the crystal structure of TcCGT1, and interpreted
structural mechanisms for its catalytic promiscuity. Moreover,
structure-guided mutagenesis was conducted to alter the C-/O-
glycosylation catalytic specificity.
HO
A
OH
OH
R
OH
R
H
8
O
HO
8
OH
HO
O
O
TcCGT1
UDP-Glc
HO
O
O
6
OH
6
OH
Results and Discussion
Apigenin (1: R=H)
Luteolin (2: R=OH)
Vitexin (1a: R=H)
Orientin (2a: R=OH)
Molecular Cloning and Functional Characterization of
TcCGT1
B
1
MS
MS2
λ =340 nm
311.08
431.11
431.1
311.1
100
Control
100
[M-H-120]−
[M-H]−
1a
1a
80
60
40
20
0
80
1a
1a
Reaction
60
40
20
We identified TcCGT1,
a
putative flavone 8-C-
glycosyltransferase from the transcriptome (BioProject accession
number PRJNA532685) of Trollius chinensis based on RNA
sequencing and bioinformatics analysis. A phylogenetic tree was
constructed to analyze the relationship of TcCGT1 with known
CGTs. As a result, TcCGT1 was grouped together with the
isoflavone 8-C-glycosyltransferase PlUGT43, though it shared
very low sequence identity (25%–30% amino acid identity) with all
reported plant CGTs (Figure S1). Similar to PlUGT43 and the
flavone 6-C-glycosyltransferase GtUF6CGT1, TcCGT1 does not
contain the conserved “DPFXL” motif for 2-hydroxyflavanone
CGTs.[8] These evidences suggested TcCGT1 may be a CGT
which could directly use flavonoids as substrates. To confirm this
presumption, we cloned the gene from T. chinensis by RT-PCR
(Table S1). The cDNA sequence of TcCGT1 (GenBank accession
number MK644229, Table S2) contains an open reading frame
(ORF) of 1452 bp encoding 483 amino acids. Recombinant
TcCGT1 was subsequently expressed in Escherichia coli and
purified by His-tag affinity chromatography (purity > 95%, Figure
S2).
Standard
341.0
[M-H-90]−
341.01
311.17
544.74
5
10
15
20
25
0
400
400
Retention time (min)
m/z
m/z
C
MS
MS2
2
447.1
327.2
327.15
[M-H-120]−
2a
447.08
100
[M-H]−
100
λ =340 nm
Control
2a
80
60
40
20
0
80
60
40
20
0
2a
2a
[M-H-90]−
357.0
357.04
Reaction
492.77
369.03
327.17
Standard
560.74
600
5
10
15
20
25
400
m/z
400
600
Retention time (min)
m/z
Figure 1. C-glycosylation of 1 and 2 catalyzed by recombinant TcCGT1. A)
TcCGT1 catalyzed the C-glycosylation of 1 and 2. B) LC/MS analysis of 1 and
the enzymatic product 1a. C) LC/MS analysis of 2 and the enzymatic product
2a. The HPLC and LC/MS parameters are given in Table S3.
The biochemical characteristics of recombinant TcCGT1 were
investigated using 1 as the acceptor and UDP-Glc as the sugar
donor. TcCGT1 showed its maximum activity at pH 8.0 (50 mM
Na2HPO4-NaH2PO4 buffer) and 30 °C, and was independent of
divalent cations (Figure S3). Kinetic analysis demonstrated that
TcCGT1 exhibited Km values of 9.0 μM and 42.3 μM for 1 and
UDP-Glc, respectively, and the corresponding kcat values were 1.1
s-1 and 0.4 s-1. The Km values for 2 and UDP-Glc were 11.8 μM
and 43.1 μM, and the corresponding kcat values were 1.2 s-1 and
0.1 s-1, respectively (Figure S4). These values indicated high
affinity of TcCGT1 toward the flavones.
To characterize the catalytic function of TcCGT1 in vitro,
apigenin (1) and luteolin (2), the proposed biosynthetic
precursors,[8e,13a] were used as sugar acceptors, and uridine 5'-
diphosphate glucose (UDP-Glc) was used as the sugar donor.
The reaction mixtures (50 mM pH 8.0 NaH2PO4-Na2HPO4; 0.5
mM UDP-Glc; 0.2 mM aglycone; 50 μg of purified recombinant
TcCGT1; 30 °C, 12 h) were analyzed by liquid chromatography
coupled with mass spectrometry (LC/MS). Heat-inactivated
enzymes (100 °C, 15 min) were used as the negative control.
Convincingly, TcCGT1 exhibited high C-glycosylation activity
toward 1 and 2 (up to 100% conversion by HPLC analysis). In the
reaction mixture of 1, a new product 1a was observed (Figure 1).
Its [M−H]- ion appeared at m/z 431, which was 162 amu greater
than 1. The MS/MS spectrum showed fragment ions at m/z 341
[M−H−90]- and m/z 311 [M−H−120]-, which were characteristic for
The Catalytic Promiscuity and Synthetic Applicability of
TcCGT1
To explore the catalytic promiscuity of TcCGT1, an acceptor
library of 114 structurally diverse substrates was tested with UDP-
Glc as the sugar donor (Figure 2, Figures S5–S72). The
substrates include flavonoids (1–36, 38–70, 84–104),
hydroxynaphthalenes (37, 71–73), lignans (74 and 75), stilbenes
This article is protected by copyright. All rights reserved.

10.1002/anie.201905505
Angewandte Chemie International Edition
RESEARCH ARTICLE
Flavones
Flavonols

Flavanones
Others
A
 


Mono‐C‐glucoside.a
Mono-C-glucoside.a Mono-C-glucoside.b
Mono‐C‐glucoside.b
C
100
Mono‐N‐glucoside
Mono-N-glucoside
Mono‐O‐glucoside
Mono-O-glucoside


Mono‐S‐glucoside
Mono-S-glucoside
80



60





40
20




0
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 40 42 82 83
B
Compound number
OH
OH
HO
OH
HO
OH
OH
OH
R
R
R
R
R
R
R
R
R
8
HO
O
O
HO
O
HO
O
HO
O
O
HO
O
O
HO
O
HO
HO
O
O
O
O
6
OH
OH
OH
1
O
OH
3
O
OH
5
O
OH
O
OH
4
O
OH
OH
2
O
R=H
R=H
10 R=H
14 R=H
14a R=Glc
15 R=H
15a R=Glc
R=H
R=H
R=H
3a R=Glc
6
R=H
6a R=Glc
4a R=Glc
5a R=Glc
10a R=Glc
1a R=Glc
2a R=Glc
OH
OH
OH
HO
OH
HO
OH
HO
OH
OH
OH
OH
HO
HO
O
OH
HO
R
O
R1
8
R
R
O
R
R
O
OH
2
HO
R2
O
O
O
O
HO
O
O
O
HO
HO
O
OH
OH
OH
O
OH
OH
6
OH
O
OH
OH
24 R=H
O
OH
O
OH O
OH
OH
O
OH O
OH
O
16 R=H
16a R=Glc
17
18
21 R=H
22 R=H
23 R=H
26
27 R=H
21a R=Glc
22a R=Glc
23a R=Glc
24a R=Glc
27a R1=Glc
27b R2=Glc
OH
R
1
R
3'
OH
1
OH
OH
HO
OH
HO
OH
HO
OH
OH
O
R1
O
HO
OH
OH
HO
RO
O
O
O
O
HO
O
O
RO
7
O
O
OH
HO
HO
Cl
5'
5
OH
R2
OH
Cl
OH
OH
O
R
OH
O
OH
OH
O
OH
O
OH
O
OH
82 R=NH2
82a R=NH-Glc
83 R=SH
29 R=H
29a R1=Glc
29b R2=Glc
Glc=β-D-glucosyl
42 R=H
42a R=Glc
28
35
36 R=H
36a R=Glc
37 R=H
40 R=H
40a R=Glc
34
37a R=Glc
83a R=S-Glc
Figure 2. Catalytic promiscuity of TcCGT1. A) Conversion rates of different substrates. Experiments were performed in triplicate (n = 3). B) Structures of part of the
substrates and corresponding glycosylated products. * indicates the products were isolated and confirmed by NMR spectroscopy, except that 1a, 2a, 36a, 82a and
83a were confirmed by comparing with reference standards. ∆ represents new compounds. Structures of other substrates are given in Figure S5. The O-glucosylation
conversion rates are shown in Figure S6.
(76 and 77), anthraquinone (78), benzophenone (79), curcumin
(80), coumarins (105 and 106), triterpenoids (107 and 108), and
simple aromatic compounds with –OH, –SH, or –NH2 groups (81–
83 and 109–114). LC/MS analysis revealed that TcCGT1 could
catalyze the glucosylation of 83 substrates (1–83, Tables S3 and
S4).
feature
a
2,4,6-trihydroxybenzophenone-like
structure
fragment.[7c,9] Thus, TcCGT1 exhibited broader substrate
promiscuity than previously reported CGTs.
In addition to C-glycosylation, TcCGT1 also showed robust
capabilities of O-glycosylation. It could catalyze the O-
glucosylation of 44 substrates (38–81), including chromone (38),
flavonoids (39–70), hydroxynaphthalenes (71–73), lignans (74
and 75), stilbenes (76 and 77), anthraquinone (78),
benzophenone (79), curcumin (80), and simple phenolics (81).
The conversion rates were >80% for 8 substrates (Figure S6).
The products were identified as O-glucosides according to the
diagnostic [M−H−162]- fragment ions in the MS/MS spectra. The
substrate promiscuity of TcCGT1 to catalyze O-glycosylation was
as broad as many versatile OGTs.[17] Interestingly, TcCGT1
exhibited both C- and O-glycosylation activities towards 17
substrates (7–10, 12, 13, 19, 20, 22–24, 27, 30–33, 36).
Surprisingly, TcCGT1 showed unprecedented substrate
promiscuity of C-glycosylation. It could catalyze 36 flavonoids of
different structural types, including flavones (1–13), flavonols (14–
25), 2-hydroxyflavanone (26), flavanones (27–33), flavanonols
(34 and 35), and dihydrochalcone (36). The products were
identified as C-glycosides according to the diagnostic fragment
ions [M−H−90]- and [M−H−120]- in the MS/MS spectra.[16] For 12
substrates (1–5, 14–16, 18, 21, 28 and 29), the conversion rates
were > 77%. It is noteworthy that TcCGT1 showed high catalytic
capabilities towards flavones, flavonols, and flavanones (up to
100%
conversion).
TcCGT1
could
also
catalyze
TcCGT1 also possessed N- and S-glycosylation activities
toward 3,4-dichloroaniline (82) and 3,4-dichlorobenzenethiol (83),
respectively. MiCGT had been reported to show N- but not S-
glycosylation activity.[7c] Thus, TcCGT1 may be the first
hydroxynaphthalene (37), which is the first example of enzymatic
C-glycosylation of this structural type. To our knowledge, most
known CGTs exhibit relatively narrow substrate promiscuity.[8]
Although two benzophenone CGTs, MiCGT and MiCGTb from M.
indica, could accept a variety of substrates, majority of them
This article is protected by copyright. All rights reserved.

10.1002/anie.201905505
Angewandte Chemie International Edition
RESEARCH ARTICLE
glycosyltransferase to catalyze all the four types of glycosylation
Glc was hydrolyzed spontaneously during crystallization. The
structure was eventually identified as a TcCGT1/UDP complex.
This phenomenon was similar to the ubiquitous situation in many
GT structures such as UrdGT2 (PDB ID: 2P6P),[10g] UGT72B1
(PDB ID: 2VCE),[10d] UGT71G1 (PDB ID: 2ACW),[10a] UGT78G1
(PDB ID: 3HBJ),[10f] and GtfD (PDB ID: 1RRV).[21]
reactions, i.e. C-, O-, N-, and S-glycosylation.
To explore the sugar donor promiscuity of TcCGT1, we tested
five other donors aside from UDP-Glc, i.e. UDP-xylose (UDP-Xyl),
UDP-galactose (UDP-Gal), UDP-arabinose (UDP-Ara), UDP-
glucuronic acid (UDP-GluA), and UDP-N-acetylglucosamine
(UDP-GlcNAc). Compounds 1 and 2 were used as substrates.
The results demonstrated that TcCGT1 could accept UDP-Xyl,
UDP-Gal, and UDP-Ara, though the conversion rates for UDP-Glc
and UDP-Xyl were relatively high (Figures S8, S73 and S74).
To fully identify structures of the products, we isolated 18 C-
glycosides (3a–6a, 10a, 14a–16a, 21a–24a, 27a, 27b, 29a, 29b,
37a, 1b) from preparative-scale enzymatic reactions. All the
products are glucosides except that 1b is a galactoside. Among
them, 10 products are new compounds (4a, 5a, 10a, 16a, 21a–
24a, 27b, and 37a). Their structures were established by HR-ESI-
MS, together with 1D and 2D NMR spectroscopic analyses
(Figures S75–S164). According to the HMBC spectra, the sugar
moieties were attached at C-8 of flavones (1a–6a, 10a and 1b)
and flavonols (14a–16a, 21a–24a), or C-3' of dihydrochalcone
(36a), indicating high regio-specificity of TcCGT1.[8a,15,18]
Interestingly, the regio-specificity was compromised for
flavanones, where TcCGT1 could generate two mono-C-
glucosides. As exemplified by 27a/27b and 29a/29b, the sugar
moieties were attached at C-8 and C-6, respectively. For all the
obtained C-glycosides, the glycosidic bonds were in the β-
configuration, according to the large coupling constants (J =
9.4−9.9 Hz) of the anomeric protons (Table S5).[7c,8a,18] Moreover,
three O-glycosylated products (40a, 42a and 72a, Figures S165–
S172) were also obtained by preparative-scale reactions.
Figure 3. Crystal structure of TcCGT1 (PDB ID: 6JTD). A) The two molecules
in the asymmetric unit of TcCGT1/UDP crystal structure. B) Superimposition of
the two TcCGT1/UDP molecules in the asymmetric unit. C) The N-terminal
domain (NTD), middle domain (MD), and C-terminal domain of TcCGT1. D) The
TcCGT1/UDP-Glc model was constructed based on superimposition of
MD/UDP in TcCGT1/UDP structure (slate) with those of A. thaliana UGT72B1
(salmon, PDB ID: 2VCE), M. truncatula UGT71G1 (yellow, PDB ID: 2ACW), and
V. vinifera VvGT1 (green, PDB ID: 2C1Z). The protein molecules are shown as
cartoons. UDP, UDP-sugar, and the sugar acceptor compounds are shown as
sticks.
Given the significant anti-inflammatory activities of T. chinensis
in traditional Chinese medicine clinical practice, we evaluated
compounds 1a, 2a, and other flavonoid 8-C-glycosides obtained
in this study for their inhibitory activities against nuclear factor-
kappa B (NF-κB) and cyclooxygenase 2 (COX-2).[19,20] A number
of C-glycosides showed similar or even higher NF-κB inhibition
activities than the corresponding substrates (Figure S173).
Among them, six compounds (1a, 1b, 2a, 14a, 27a and 29b)
inhibited NF-κB by > 33% at 10 µM (the positive control MG132
showed an inhibition rate of 55% at 10 µM). Moreover, seven
compounds (2a, 14a, 16a, 23a, 24a, 29b and 37a) exhibited
potent COX-2 inhibitory activities (> 80% at 10 µM).
The crystal structure, in the asymmetric unit, contains two
TcCGT1/UDP molecules. The two molecules are highly similar to
each other, with root-mean-square deviation (RMSD) of 0.328 Å
for all atoms of 471 residues. Remarkable conformational
difference was observed only between residues 78–100, which
participate in the formation of the putative entry channel for the
sugar acceptor and thus may represent intrinsic flexibility of this
region needed in entry/exit of the substrate/product (Figure 3A
and 3B). TcCGT1 exhibits a typical GT-B fold structure consisting
of two lobes (Figure 3C). The N-terminal domain (NTD, residues
1–251) and C-terminal domain (CTD, residues 461-483) forms
one lobe (NC) responsible mainly for sugar acceptor binding,
while the middle domain (MD, residues 252–460) is responsible
for UDP-sugar binding. Both lobes adopt the “Rossmann-like
Crystal Structure of TcCGT1/UDP Complex
To understand structural basis for the catalytic mechanisms
and substrate promiscuity of TcCGT1, we obtained its crystal
structure at 1.85Å resolution (PDB ID: 6JTD), in the presence of
UDP-Glc (Table S6). Although the electron density of nearly all
the protein residues were well defined, the glucose moiety was
invisible in the crystal structure (Figures S174 and S175),
probably because TcCGT1 bound UDP endogenously, or UDP-
(β/α/β)
fold”,
a
conserved
domain
in
plant
glycosyltransferases,[10,22] though they do not resemble each
other. The two lobes pack tightly to form a deep cleft which acts
as the binding site of the sugar acceptor substrate. Sequence
This article is protected by copyright. All rights reserved.

10.1002/anie.201905505
Angewandte Chemie International Edition
RESEARCH ARTICLE
Figure 4. Molecular docking and mutagenesis studies on TcCGT1. A-D) Predicted binding modes of selected substrates and their glycosylated products to
TcCGT1/UDP-Glc and TcCGT1/UDP by docking. Apigenin (1, A, left), vitexin (1a, A, right), wogonin (40, B, left), wogonin 7-O-glucoside (40a, B, right), luteolin (2,
C, left), orientin (2a, C, right), licoflavone C (42, D, left) and its 7-O-glucoside (42a, D, right). The protein molecules are shown as cartoons. The key residues and
substrate molecules are shown as sticks. Labels with and without brackets indicate predicted (by molecular docking) and experimentally determined models of the
small molecules, respectively. Dashes indicate hydrogen bonds. E) Catalytic activities of wild-type (WT) TcCGT1 and its mutants (H24A and E396A) towards
substrates 1 and 2. F) Catalytic activities of wild-type (WT) TcCGT1 and its mutants (H24A and E396A) towards substrates 40 and 42. N.D., products not detected
(Figure S178).
alignment (Figure S176) and three-dimensional structure
superimposition (Figure 3D and Figure S177) revealed that
although TcCGT1 is a robust CGT, it does not show significant
similarity either in primary sequence or in tertiary structure to the
bacterial CGTs UrdGT2 (2P6P, sequence identity about 16% and
RMSD about 13 Å) or SsfS6 (4G2T, sequence identity about 16%
and RMSD about 17 Å).[10g,10h,23] Interestingly, TcCGT1 shares
high homology with other plant glycosyltransferases such as
UGT72B1 (2VCE, sequence identity about 40% and RMSD about
1.243Å) which is a bifunctional N- and O-glucosyltransferase
(NGT and OGT) from Arabidopsis thaliana,[10d] UGT71G1 (2ACW,
sequence identity about 27% and RMSD about 1.661 Å) which is
the glucose or analogue moieties observed in these structures
(Figure 3D). We then built the models for TcCGT1/UDP-
Glc/apigenin (1), TcCGT1/UDP-Glc/luteolin (2), TcCGT1/UDP-
Glc/wogonin (40) and TcCGT1/UDP-Glc/licoflavone
C (42)
through computer aided molecular docking of the substrates 1
and 2 (two optimal C-glycosylation substrates of TcCGT1), 40 and
42 (two representative O-glycosylation substrates of TcCGT1)
into the TcCGT1/UDP-Glc structure. We also built the
TcCGT1/UDP/1a, TcCGT1/UDP/2a, TcCGT1/UDP/40a and
TcCGT1/UDP/42a models by docking products 1a, 2a, 40a and
42a into the TcCGT1/UDP structure (Figure 4).
a
multifunctional triterpene/flavonoid OGT from Medicago
truncatula,[10a] and VvGT1 (2C1Z, sequence identity about 25%
and RMSD about 1.854 Å) which is flavonoid 3-O-
The molecular docking provided interesting information that
may reveal catalytic mechanisms for the C- and O-glycosylation
activities of TcCGT1. Substrate 1 or 2 binds to TcCGT1 in a
direction in which C-8 is placed in the most proximity to C-1' of the
glucose moiety of UDP-Glc, which could explain why the sugar is
transferred to C-8 position of the substrate (8-C-glycosylation)
(Figure 4A and 4C). In contrast, substrate 40 or 42 bound to
TcCGT1 is placed in a different direction almost perpendicular to
that of 1 or 2. The binding pose of 40 and 42 clearly locates 7-OH
close to C-1' of the sugar, which could explain why TcCGT1
catalyzes transfer of glucose to 7-OH of the substrates (7-O-
glycosylation) (Figure 4B and 4D).
a
glycosyltransferase from Vitis vinifera.[10b] These structural
comparisons indicate the catalytic mechanism of TcCGT1 may be
different from bacterial CGTs.
Catalytic Mechanisms for C- and O-Glycosylation of TcCGT1
TcCGT1 exhibits both C- and O-glycosylation activities towards
various flavonoids. In order to investigate the catalytic
mechanisms of TcCGT1, we tried to obtain the complex structures
of TcCGT1 bound by a substrate (such as 1) or a product (such
as 1a). However, after a lot of trials including co-crystallization and
soaking experiments, we were unable to obtain such a complex.
Fortunately, since TcCGT1 shares high structural similarity to
UGT72B1 (2VCE), UGT71G1 (2ACW) and VvGT1 (2C1Z), we
could readily model the glucose moiety of UDP-Glc according to
It had been proposed that deprotonation of a hydroxyl group
activated by a histidine (His) residue is the key step to initiate GT-
mediated glycosylation.[10a-c,10e,10f] In TcCGT1, this key residue
was mapped to His24 (H24). However, in the TcCGT1/UDP-
Glc/substrate docking models, H24 is located far away from any
This article is protected by copyright. All rights reserved.

10.1002/anie.201905505
Angewandte Chemie International Edition
RESEARCH ARTICLE
Figure 5. Comparative structural analysis and structure-guided mutagenesis of TcCGT1. A) The spacious sugar acceptor binding pocket of TcCGT1. B) The
crowded (compared to TcCGT1) sugar acceptor binding pocket of OGTs, including UGT72B1 (salmon), UGT71G1 (yellow) and VvGT1 (green). C) Assuming
apigenin binds to UGT72B1 (salmon) or UGT71G1 (yellow) in the same way as it binds to TcCGT1, UGT72B1 Tyr315 or UGT71G1 Met286 would clash with the
substrate. D) Assuming apigenin binds to VvGT1 (green) in the same way as it binds to TcCGT1, VvGT1 Val281 would clash with the compound. The protein
molecules are shown as cartoons. The key residues and substrate molecules are shown as sticks. Labels with and without brackets indicate predicted (by molecular
docking) and experimentally determined models of the small molecules, respectively. E–H) Catalytic activities of wild-type (WT) TcCGT1 and its mutants towards
substrates 1, 2, 40 and 42 (Figures S180–S183). 1a: vitexin (apigenin 8-C-glucoside), 1b: apigenin 7-O-glucoside, 1c: apigenin 4'-O-glucoside, 1d: apigenin 7,4'-
di-O-glucoside. 2a: orientin (luteolin 8-C-glucoside). 40a: wogonin 7-O-glucoside. 42a: licoflavone C 7-O-glucoside.
hydroxyl group of the sugar acceptor (including 7-OH), which did
not well support the previous hypothesis. Interestingly, ligand
preparation process with the Schördinger® GLIDE software
package indicated that 7-OH, but not 5-OH of 1, 2, 40 or 42 would
undergo spontaneous deprotonation in physiological condition.
Therefore, although H24 may still act as a proton acceptor to
facilitate 7-OH deprotonation, it does not seem to be
indispensable. Further site-directed mutagenesis studies were
conducted to test this hypothesis. Indeed, mutation of H24 to
alanine (H24A) did weaken, but did not abolish the C- or O-
glycosylation activities of TcCGT1 (Figure 4E, Figure 4F, and
Figure S178). On the other hand, docking of the glycosylated
products, i.e. 1a, 2a, 40a or 42a into the TcCGT1/UDP structure
indicated that H24 may stabilize the products through hydrogen
bonding and thus facilitate the catalytic process. Taken together,
the docking studies indicated a new model for the catalytic
mechanism of TcCGT1 which includes the following steps: 1)
spontaneous deprotonation of 7-OH of the flavone substrate
results in negative charge on 7-O or C-8 atom (due to electron
rearrangement on the aryl ring), and H24 side-chain imidazole
may stabilize the deprotonated substrate; 2) the negatively
charged 7-O or C-8 atom, whichever is physically closer, would
attack C-1' of the sugar to fulfill the reaction; 3) H24 side-chain
imidazole interacts with and stabilizes the products through
hydrogen bonding with sugar hydroxyls of the products to facilitate
the reaction (Scheme S2).
Moreover, in the docking models of TcCGT1/UDP-
Glc/substrate, Glu396 (E396) was very close to the sugar to form
hydrogen bonds with the sugar hydroxyls. This residue was
considered a key residue to stabilize the donor substrate and
position the sugar in the optimal orientation for the glycosylation
reaction. Mutation of E396 to alanine (E396A) indeed drastically
decreased or even abolished the catalytic activity of TcCGT1
(Figure 4E, Figure 4F, and Figure S178).
Based on the model presented above, E396 plays an important
role to stabilize and orient the UDP-Glc sugar. H24 acts to
stabilize both the deprotonated substrate and the product sugar,
though it is not indispensable for the glycosylation activity. The
substrate flavonoids may undergo spontaneous deprotonation to
initiate the glycosylation reaction. Most importantly, both C- and
O-glycosylation activities of TcCGT1 share the common catalytic
mechanism. Whether TcCGT1 acts as CGT or OGT is determined
by binding pose of the substrate, i.e. which of the negatively
charged 8-C- or 7-O- is located closer to C-1' of the sugar.
The Substrate Promiscuity is Enabled by the Spacious
Binding Pocket
To further explore structural basis for the substrate promiscuity
of TcCGT1, we analyzed structure of the active sites at the
substrate binding pocket. The methoxy or dimethylallyl group at
C-8 of 40 and 42 makes it impossible for these substrates to bind
This article is protected by copyright. All rights reserved.

10.1002/anie.201905505
Angewandte Chemie International Edition
RESEARCH ARTICLE
to TcCGT1 in the same mode as 1 or 2 does, due to steric
hindrance with the sugar of UDP-Glc or the side-chain of H24
(Figure S179). However, 1 and 2 could presumably bind to
TcCGT1 in the same mode as 40 and 42 do. Comparison of the
active site pocket between TcCGT1 with those of UGT72B1,
UGT71G1 and VvGT1 revealed that TcCGT1 has the most
spacious sugar acceptor binding pocket. This spacious pocket
can readily accommodate the 2-phenyl moieties of 1 or 2 (Figure
5A), allows 1 or 2 to bind more deeply into the enzyme than 40 or
42 does (Figure 4 and Figure S179), and thus enables the C-
glycosylation activity. In UGT72B1, UGT71G1 and VvGT1,
however, the same space is occupied by Tyr315, Met286 and
Val281, respectively (Figure 5B). Taken together, the docking
studies indicate the spacious sugar acceptor binding pocket of
TcCGT1 enables the unique binding mode of 1 or 2 to the enzyme
to enable the C-glycosylation activity. The spacious binding
pocket also explains why TcCGT1 shows broad substrate
promiscuity.
Conclusion
In summary, we characterized
a
promiscuous C-
glycosyltransferase TcCGT1 from the medicinal plant T. chinensis.
TcCGT1 represents the first plant CGT with a crystal structure and
the first flavone 8-C-glycosyltransferase. It could region-
specifically catalyze 8-C-glycosylation of flavones, flavonols, and
other types of flavonoids. TcCGT1 is also the first
glycosyltransferase that can catalyze C-, O-, N-, and S-
glycosylation reactions. The 1.85 Å resolution crystal structure of
TcCGT1/UDP complex, molecular docking, and site-directed
mutagenesis revealed a new model of catalytic mechanism of
TcCGT1. The substrate flavonoids undergo spontaneous
deprotonation to initiate the glycosylation reaction, and the
residues H24 and E396 play an important role in stabilizing and
orienting the small molecules. The broad substrate promiscuity of
TcCGT1 is enabled by the spacious binding pocket. The
selectivity between C- or O-glycosylation activities is determined
by the binding pose of the substrate at the pocket, and the
mutations at I94E and G284K installed the O-glycosylation activity
of TcCGT1 while abolishing the C-glycosylation activity. This
study provides a basis to design effective biocatalysts for efficient
and directed biosynthesis of important bioactive flavonoid C-
glycosides for drug discovery.
Structure-Guided Engineering of the C- or O-Glycosylation
Selectivity of TcCGT1
In order to explore the determination between C- and O-
glycosylation activities of TcCGT1, we tended to conduct site-
directed mutagenesis of residues around the spacious sugar
acceptor binding pocket to shrink it. The shrinked pocket would
force typical CGT substrates of TcCGT1 (such as 1 and 2) to bind
in the way as 40 and 42 do, thus leading to O-glycosylation. We
inspected the binding pocket of typical OGTs, and found the bulky
side-chains of methionine (Met286 in UGT71G1), tyrosine
(Tyr315 in UGT72B1), or valine (Val281 in VvGT1) would
presumably block the binding of 1 or 2 in the CGT substrate way
(Figure 5C and 5D). In TcCGT1, the corresponding residue is
Gly284 (G284). We then mutated Gly284 to an amino acid with a
bulky side-chain, such as Phe (G284F), Gln (G284Q), Tyr
(G284Y) and Lys (G284K). We also intended to build a salt-bridge
between residues I94 and G284 to occupy the pocket. Thus we
prepared G284F, G284Q, G284Y, G284K and I94E single
mutants or I94E/G284K double mutant of TcCGT1 to see whether
these modifications can convert the activity of TcCGT1 towards 1
or 2 from CGT to OGT. Consistent with our hypothesis, all the
mutations drastically decreased the CGT activity of TcCGT1
towards 1 or 2, and some mutants showed significant O-
glycosylation activities (Figure 5E, Figure 5F, and Figures S180–
S183). In the case of the natural OGT substrate 40, these
mutations did not alter the O-glycosylation activity or even
enhanced the O-glycosylation activity (Figure 5G). For substrate
42, however, the mutations partially inhibited the O-glycosylation
activity (Figure 5H). This observation was consistent with our
model. The isoprenyl group at C-8 of 42 is bulkier than the
methoxyl group of 40. Substrate 42 protrudes more deeply into
the binding pocket than the latter, and was therefore affected
more by the mutations.
Acknowledgements
The authors thank Dr. Kuan Chen at Peking University for
providing the authentic reference standards 1b–1d, and Dr. Kai Li
for collecting the plant samples. This work was supported by
National Natural Science Foundation of China (81725023,
81891010/81891011, 31270769), Beijing Natural Science
Foundation (JQ18027), and Young Elite Scientists Sponsorship
Program by China Association for Science and Technology
(2016QNRC001).
Conflict of interest
The authors declare no conflict of interest.
Keywords: glycosylation • enzyme catalysis •
glycosyltransferase • crystal structure • catalytic mechanisms
[1]
a) C. J. Thibodeaux, C. E. Melancon, H. W. Liu, Nature 2007, 446, 1008-
1016; b) C. J. Thibodeaux, C. E. Melancon III, H. W. Liu, Angew. Chem.
Int. Ed. 2008, 47, 9814-9859; c) T. Bililign, B. R. Griffith, J. S. Thorson,
Nat. Prod. Rep. 2005, 22, 742-760.
[2]
[3]
P. G. Hultin, Curr. Top. Med. Chem. 2005, 5, 1299-1331.
a) J. Xiao, E. Capanoglu, A. R. Jassbi, A. Miron, Crit. Rev. Food Sci. Nutr.
2016, 56, S29-S45; b) O. Talhi, A. M. S. Silva, Curr. Org. Chem. 2012,
16, 859-896.
[4]
[5]
a) K. Y. Lam, A. P. K. Ling, R. Y. Koh, Y. P. Wong, Y. H. Say, Adv.
Pharmacol. Sci. 2016, Article ID 4104595; b) M. He, J. W. Min, W. L.
Kong, X. H. He, J. X. Li, B. W. Peng, Fitoterapia 2016, 115, 74-85.
S. Sato, T. Akiya, H. Nishizawa, T. Suzuki, Carbohydr. Res. 2006, 341,
964-970.
This article is protected by copyright. All rights reserved.

10.1002/anie.201905505
Angewandte Chemie International Edition
RESEARCH ARTICLE
[6]
a) H. Satoh, S. Manabe, Chem. Soc. Rev. 2013, 42, 4297-4309; b) Y.
Yang, B. Yu, Chem. Rev. 2017, 117, 12281-12356.
[7]
a) E. K. Lim, Chem. Eur. J. 2005, 11, 5486-5494; b) C. Durr, D.
Hoffmeister, S. E. Wohlert, K. Ichinose, M. Weber, U. von Mulert, J. S.
Thorson, A. Bechthold, Angew. Chem. Int. Ed. 2004, 43, 2962-2965; c)
D. W. Chen, R. D. Chen, R. S. Wang, J. H. Li, K. B. Xie, C. C. Bian, L. L.
Sun, X. L. Zhang, J. M. Liu, L. Yang, F. Ye, X. M. Yu, J. G. Dai, Angew.
Chem. Int. Ed. 2015, 54, 12678-12682.
[8]
a) M. Brazier-Hicks, K. M. Evans, M. C. Gershater, H. Puschmann, P. G.
Steel, R. Edwards, J. Biol. Chem. 2009, 284, 17926-17934; b) M. L.
Hamilton, S. P. Kuate, M. Brazier-Hicks, J. C. Caulfield, R. Rose, R.
Edwards, B. Torto, J. A. Pickett, A. M. Hooper, Phytochemistry 2012, 84,
169-176; c) M. L. F. Ferreyra, E. Rodriguez, M. I. Casas, G. Labadie, E.
Grotewold, P. Casati, J. Biol. Chem. 2013, 288, 31678-31688; d) Y.
Nagatomo, S. Usui, T. Ito, A. Kato, M. Shimosaka, G. Taguchi, Plant J.
2014, 80, 437-448; e) N. Sasaki, Y. Nishizaki, E. Yamada, F. Tatsuzawa,
T. Nakatsuka, H. Takahashi, M. Nishihara, FEBS Lett. 2015, 589, 182-
187; f) Y. Hirade, N. Kotoku, K. Terasaka, Y. Saijo-Hamano, A. Fukumoto,
H. Mizukami, FEBS Lett. 2015, 589, 1778-1786; g) B. Hao, J. C. Caulfield,
M. L. Hamilton, J. A. Pickett, C. A. O. Midega, Z. R. Khan, J. Wang, A.
M. Hooper, Phytochemistry 2016, 125, 73-87; h) X. Wang, C. F. Li, C.
Zhou, J. Li, Y. S. Zhang, Plant J. 2017, 90, 535-546; i) T. Ito, S. Fujimoto,
F. Suito, M. Shimosaka, G. Taguchi, Plant J. 2017, 91, 187-198.
D. W. Chen, L. L. Sun, R. D. Chen, K. B. Xie, L. Yang, J. G. Dai, Chem.
Eur. J. 2016, 22, 5873-5877.
[9]
[10] a) H. Shao, X. Z. He, L. Achnine, J. W. Blount, R. A. Dixon, X. Q. Wang,
Plant Cell 2005, 17, 3141-3154; b) W. Offen, C. Martinez-Fleites, M.
Yang, E. Kiat-Lim, B. G. Davis, C. A. Tarling, C. M. Ford, D. J. Bowles,
G. J. Davies, EMBO J. 2006, 25, 1396-1405; c) L. Li, L. V. Modolo, L. L.
Escamilia-Trevino, L. Achnine, R. A. Dixon, X. Wang, J. Mol. Biol. 2007,
370, 951-963; d) M. Brazier-Hicks, W. A. Offen, M. C. Gershater, T. J.
Revett, E.-K. Lim, D. J. Bowles, G. J. Davies, R. Edwards, Proc. Natl.
Acad. Sci. U. S. A. 2007, 104, 20238-20243; e) T. Hiromoto, E. Honjo, N.
Noda, T. Tamada, K. Kazuma, M. Suzuki, M. Blaber, R. Kuroki, Protein
Sci. 2015, 24, 395-407; f) L. V. Modolo, L. Li, H. Pan, J. W. Blount, R. A.
Dixon, X. Wang, J. Mol. Biol. 2009, 392, 1292-1302. g) M. Mittler, A.
Bechthold, G. E. Schulz, J. Mol. Biol. 2007, 372, 67-76; h) F. B. Wang,
M. Q. Zhou, S. Singh, R. M. Yennamalli, C. A. Bingman, J. S. Thorson,
G. N. Phillips, Proteins 2013, 81, 1277-1282; i) D. Foshag, C. Campbell,
P. D. Pawelek, BBA-Proteins Proteom. 2014, 1844, 1619-1630.
[11] D. W. Chen, S. Fan, R. D. Chen, K. B. Xie, S. Yin, L. L. Sun, J. M. Liu, L.
Yang, J. Q. Kong, Z. Y. Yang, J. G. Dai, ACS Catal. 2018, 8, 4917-4927.
[12] A. Gutmann, B. Nidetzky, Angew. Chem. Int. Ed. 2012, 51, 12879-12883.
[13] a) M. Yuan, R. F. Wang, X. W. Wu, Y. N. An, X. W. Yang, Chin. J. Nat.
Med. 2013, 11, 449-455; b) E. Witkowska-Banaszczak, Phytother. Res.
2015, 29, 475-500.
[14] a) L. Z. Wu, X. P. Zhang, X. D. Xu, Q. X. Zheng, J. S. Yang, W. L. Ding,
J. Pharm. Biomed. Anal. 2013, 75, 55-63; b) Z. L. Song, Y. Hashi, H. Y.
Sun, Y. Liang, Y. X. Lan, H. Wang, S. Z. Chen, Fitoterapia 2013, 91, 272-
279.
[15] a) J. X. Wei, D. Y. Li, Z. L. Li, Phytochem. Lett. 2018, 25, 156-162; b) R.
Yan, Y. Cui, B. Deng, J. Bi, G. Zhang, J. Nat. Med. 2019, 73, 297-302.
[16] X. Qiao, W. N. He, C. Xiang, J. Han, L. J. Wu, D. A. Guo, M. Ye,
Phytochem. Anal. 2011, 22, 475-483.
[17] a) K. Xie, R. Chen, J. Li, R. Wang, D. Chen, X. Dou, J. Dai, Org. Lett.
2014, 16, 4874-4877; b) C. S. Zhang, B. R. Griffith, Q. Fu, C. Albermann,
X. Fu, I. K. Lee, L. J. Li, J. S. Thorson, Science 2006, 313, 1291-1294.
[18] N. Krafczyk, M. A. Glomb, J. Agr. Food Chem. 2008, 56, 3368-3376.
[19] J. Napetschnig, H. Wu, Annu. Rev. Biophys. 2013, 42, 443-468.
[20] M. E. Turini, R. N. DuBois, Annu. Rev. Med. 2002, 53, 35-57.
[21] A. M. Mulichak, W. Lu, H. C. Losey, C. T. Walsh, R. M. Garavito,
Biochemistry 2004, 43, 5170-5180.
[22] C. Breton, S. Fournel-Gigleux, M. M. Palcic, Curr. Opin. Struct. Biol. 2012,
22, 540-549.
[23] X. Robert, P. Gouet, Nucleic Acids Res. 2014, 42, W320-W324.
This article is protected by copyright. All rights reserved.

10.1002/anie.201905505
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents (Please choose one layout)
RESEARCH ARTICLE
Jun-Bin He,† Peng Zhao,† Zhi-Min Hu,
Shuang Liu, Yi Kuang, Meng Zhang, Bin
Li, Cai-Hong Yun,* Xue Qiao,* and Min
Ye*
Page No. – Page No.
Molecular Characterization and
Structural Basis of a Promiscuous C-
Glycosyltransferase from Trollius
chinensis
A promiscuous C-glycosyltransferase TcCGT1 was highlighted. TcCGT1 represents
the first flavone 8-C-glycosyltransferase which exhibits robust substrate promiscuity
toward different types of flavonoids. TcCGT1 is also the first plant CGT with a
crystal structure. This work provides a basis for protein engineering to design
efficient glycosylation biocatalysts for drug discovery.
This article is protected by copyright. All rights reserved.