Efficient O-Glycosylation of Triterpenes Enabled by Protein Engineering of Plant Glycosyltransferase UGT74AC1

Article abstract of DOI:10.1021/acscatal.9b05232

Triterpene O-glycosylation has attracted significant interest from the pharmaceutical industry as a valuable means for drug design and development. Plant glycosyltransferases, which catalyze this glycosylation reaction, play a key step in preparing structure diverse and valuable triterpene glycosides. However, this class of enzymes usually suffers from low catalytic efficiency. To address this problem, triterpene glycosyltransferase UGT74AC1 from Siraitia grosvenorii was chosen and its crystal structure was solved and employed as the molecular basis to implement directed evolution and sequence/structure-based engineering. Several resultant uridine diphosphate (UDP) glycosyltransferases (UGTs) variants exhibit a 10²- to 10⁴-fold improvement in catalytic efficiency for triterpene glycosylation. Especially, one variant exhibited up to 4.17 × 10⁴-fold increase in catalytic efficiency toward mogrol and 1.53 × 10⁴-fold increase to UDP-glucose, respectively. Moreover, the mutants also displayed extended substrate promiscuity compared with wild-type enzyme and conserved regioselectivity. Based on the results of molecular docking and molecular dynamics simulations, it was proposed that the improved enzymatic activity and substrate promiscuity were likely owing to the stable hydrophobic interactions and favorite conformations between the enzyme and substrates. This work has also laid a foundation for the engineering of other plant UGTs for their practical application to the synthesis of valuable triterpene saponins.

Full text of DOI:10.1021/acscatal.9b05232

Subscriber access provided by La Trobe University Library
Article
Efficient O-glycosylation of Triterpenes Enabled by Protein
Engineering of Plant Glycosyltransferase UGT74AC1
Jiao Li, Jiangang Yang, Shicheng Mu, Na Shang, Cui Liu, Yueming Zhu,
Yi Cai, Pi Liu, Jianping Lin, Weidong Liu, Yuanxia Sun, and Yanhe Ma
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.9b05232 • Publication Date (Web): 21 Feb 2020
Downloaded from pubs.acs.org on February 24, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
Efficient O-glycosylation of Triterpenes Enabled by Protein
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
Engineering of Plant Glycosyltransferase UGT74AC1
Jiao Li1,2‡, Jiangang Yang , Shicheng Mu , Na Shang , Cui Liu , Yueming Zhu , Yi Cai , Pi Liu ,
1‡
1
1
1
1
1
1
3
4
1
1
1,2
1
Jianping Lin *, Weidong Liu *, Yuanxia Sun * and Yanhe Ma
1
5
6
7
National Engineering Laboratory for Industrial Enzymes, Tianjin Institute of Industrial
Biotechnology, Chinese Academy of Sciences, 32 Xi Qi Dao, Tianjin Airport Economic Area,
Tianjin 300308, China
2
8
University of Chinese Academy of Sciences, Beijing 100049, China
9
0
1
2
3
1
1
1
1
14
1
1
5
6
17
1
8
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
9
0
ABSTRACT: Triterpene O-glycosylation has attracted significant interest from the
pharmaceutical industry as a valuable means for drug design and development. Plant
glycosyltransferases, which catalyze this glycosylation reaction, play the key step in preparing
structure diverse and valuable triterpene glycosides. However, this class of enzymes usually
suffered from low catalytic efficiency. To address this problem, triterpene glycosyltransferase
UGT74AC1 from Siraitia grosvenorii was chosen and its crystal structure was solved and
employed as the molecular basis to implement directed evolution and sequence/structure-based
21
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
2
3
24
2
2
2
5
6
7
2
4
engineering. Several resultant UGT variants exhibit a 10 - to 10 -fold improvement in catalytic
4
efficiency for triterpene glycosylation. Specially, one variant exhibited up to 4.17×10 -fold
4
28
increase in catalytic efficiency towards mogrol and 1.53×10 -fold increase to UDP-glucose,
2
3
9
0
respectively. Moreover, the mutants also displayed extended substrate promiscuity compared
with wild type enzyme and conserved regioselectivity. Based on the results of molecular docking
and molecular dynamics simulations, it was proposed that the improved enzymatic activity and
substrate promiscuity were likely owing to the stable hydrophobic interactions and favorite
conformations between the enzyme and substrates. This work has also laid a foundation for the
engineering of other plant UGTs for their practical application to the synthesis of valuable
triterpene saponins.
31
3
3
2
3
34
3
5
36
KEYWORDS: biocatalysis · catalytic efficiency · protein engineering · plant glycosyltransferase
3
3
7
8
· triterpene saponins
1. INTRODUCTION
39
Glycosylation is a key step in the synthesis of complex and structure diverse glycosides.
Chemical glycosylation usually requires tedious protection and deprotection steps and suffers
4
0
ACS Paragon Plus Environment
2

Page 3 of 34
ACS Catalysis
1
2
3
1
4
4
1
2
from low specificities and yields. Enzymatic glycosylation, on the other hand, exhibit high
efficiency and stereo/regiospecificity to create glycosidic linkages rapidly in a direct and an
environmental benign manner and are thus particularly attractive for the synthesis of glycosides.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
43
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
4
4
4
4
4
5
6
7
Terpenoids represent a large class of natural products. It has been disclosed that glycosylation
3
,4
modification of terpenoids can change their pharmacological activity and other properties, such
5
6
as taste and bioavailability as well, which has motivated considerable interest in terpene
glycosylation for drug development in the pharmaceutical industry.7
4
4
8
9
Glycosyltransferases (GTs) (EC 2.4.x.y) catalyze the transfer of glycosyl moiety from an
activated nucleotide sugar to a variety of acceptor biomolecules. These enzymes can be classified
into 110 families based on amino acid sequence similarities collected in the Carbohydrate Active
50
8
5
5
1
2
Enzyme database (CAZy, http://www.cazy.org/). The CAZy GT1 family often referred to as
9
UDP glycosyltransferases (UGTs), which accept UDP sugars as sugar donor, are widely
53
detected in the synthetic pathway of natural glycosides including terpene glycosides. Plant UGTs
which are belong to GT1 family displayed excellent properties (e.g. high regioselectivity and
substrate specificity)10 and have attracted substantial attention due to their great potential
54
55
56
57
58
59
2
11
application in biotechnology. Until now, lots of plant UGT genes from Arabidopsis thaliana,
Medicago truncatul,12 Glycyrrhiza uralensis,13 Camellia sinensis,14 Panax ginseng,15 Siraitia
grosvenorii,16,17 Glycine max, Carthamus tinctorius and so on were annotated in databases.
18
19
However, only a few plant UGTs have been characterised to be responsible for triterpene
2
0
60
glycosylation. Furthermore, the majority of these UGTs have low catalytic activity, narrow
substrate scope and difficult soluble heterologous expression, which hinder their practical
application.21, 22 Therefore, it is necessary to understand the catalytic mechanism of plant UGTs
furtherly and engineer them to improve their property and applicability.
6
6
1
2
63
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
6
4
5
Crystal structures not only provide structural basis for understanding enzyme-substrate
interaction but also lay the foundation for protein engineering. With the rapid development of
protein crystallization techniques, incresing number of GTs structure has been solved and
disclosed in PDB database. And three main kinds of structural topologies of GTs (termed as GT-
66
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
6
6
7
7
7
7
8
9
0
1
2
7
A, GT-B and GT-C folds) were identified until now. In GT1 family, there are 35 UGT crystal
structures included (statistics in CAZy), among which 13 plant UGTs structures have been
solved, including UGT74F2, UGT72B1 and UGT89C1 from A. thaliana,23-25 Os79 from Oryza
2
6
27,28
29
sativa, UGT78K6 and Ct3GT-A from C. ternatea,
VvGT1 from V. vinifera, UGT78G1,
UGT85H2 and UGT71G1 from M. truncatula,30-32 UGT76G1 from Stevia rebaudiana,33
3
4
35
73
TcCGT1 from Trollius chinensis, and PtUGT1 from P. tinctorium. All of these UGTs contain
a GT-B fold like others in GT1, consisting of two β/α/β Rossmann-like domains. Among those
UGT structures in database, there was only one plant UGT (MtUGT71G1) except for UGT51
from Saccharomyces cerevisiae (ScUGT51), exhibiting catalytic glycosylation of
triterpenoids.32,36 Limited crystal structures restricted the detailed understanding of structure–
function relationships and glycosylation mechanisms between UGTs and triterpenes.
7
7
4
5
76
7
7
8
7
7
8
8
8
9
0
1
2
Nowadays, several protein engineering strategies are available to increase the catalytic
activity and regioselectivity of UGTs. For example, directed evolution of UGT OleD from
Streptomyces antibioticus was performed using a simple high-throughput screen method based
on a fluorescent surrogate acceptor substrate.37 The triple mutant (P67T/S132F/A242V)
displayed marked improvement in catalytic efficiency and substrate promiscuity.38 The
ScUGT51 has also been engineered and its catalytic efficiency to protopanoxadiol increased
83
8
8
4
5
3
9
about 1800-fold after nine rounds of iterative saturation mutagenesis. However, only UGTPg45
86
from Panax ginseng has been engineered using direct evolution method, resulting in a minimal
ACS Paragon Plus Environment
4

Page 5 of 34
ACS Catalysis
1
2
3
8
8
7
8
(1.5-fold) increase in catalytic efficiency to protopanoxadiol.40 Therefore, dramatically
enhancing the catalytic efficiency of plant UGTs remains a challenge probably due to the lack of
sufficient molecular templates for their structural modelling and in-depth understanding of their
reaction mechanisms especially concerning the enzyme-substrate binding and interaction.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
89
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
9
0
91
92
93
94
95
We recently reported that UGT74AC1 from Siraitia grosvenorii (SgUGT74AC1) catalyzed
the high regiospecific glycosylation of 3-OH of mogrol (1a) (Figure 1) and had good soluble
expression in Escherichia coli BL21(DE3). It showed high application in preparing triterpene
glycosides such as Rh2 which possesses versatile anti-tumour and anti-inflammatory activities.41-
4
3
17
However, it showed rather weak enzymatic activity and conversion rates. In this research, we
96
analyzed the crystal structures of SgUGT74AC1 and on the basis of this information, this
enzyme was engineered to address the challenges in the O-glycosylation of triterpenes using 1a
as the model substrate. Remarkable increase in catalytic efficiency to 1a and high substrate
promiscuity were observed for the engineered mutants. Crystal structural studies and molecular
dynamics (MD) simulations of wild type (WT) and mutants were also employed to reveal their
molecular basis in glycosylation.
9
7
8
9
99
00
01
1
1
1
02
2. RESULTS AND DISCUSSION
1
1
1
03
04
05
2.1 Overall Crystal Structure of UGT74AC1. Herein, to better understand the molecular
basis of SgUGT74AC1 in glycosylation of 1a, the crystal structure of SgUGT74AC1 (PDB ID:
6L90) and its complex with UDP-glucose (UDPG) (PDB ID: 6L8Z) were solved by X-ray
crystallography at a resolution of 2.02 and 2.10 Å, respectively (Table S1 and Figure S1). As
106
7
1
1
07
08
shown in Figure 1A, the structure of SgUGT74AC1 displays the conserved GT-B fold which
features the N- (Met1–Lys226) and C-terminals (Cys259–Thr454) β/α/β Rossmann domains and
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
09
10
a less conserved connecting loop region (Thr227-Val258). Moreover, the C-terminal domain
consists of ten α-helices and six β-strands and possesses the highly conserved plant secondary
product glycosyltransferase (PSPG) motif (Trp330-Gln373), which determines the recognition
and binding of the UDP-sugar donor (Figure S2). The UDPG which is located across two helices,
C5 and C6, is perpendicular to six β-sheets of the C-terminal domain and interacts with PSPG
through the formation of several hydrogen bonds (Figure S3). Therefore, the residues around
these regions can be engineered to increase sugar specificity and catalytic activity.
111
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
12
13
114
115
A
B
Asp 111
N-domain
His 18
Asp111
C
-
O
O
His18
NH
:N
H
O
O
Mogrol
UDPG Mogrol
HO
UDPG
Mogrol
O
UDP
UDPG
C-domain
OH
OH
C
0.1
.08
.06
HO
HO
HO
HO
0
0
UGT74AC1
H
H
0.04
0.02
UDP-Glc
OH
3
O
3
HO
O
HO
HO
OH
0
Mogrol (1a)
Mogroside IE (1c)
WT
H18A
D111A
H18A/D111A
1
16
117
118
119
120
121
122
123
124
125
Figure 1. (A) Crystal structures of UGT74AC1 with UDPG (blue sticks), and mogrol (1a) was docked into the
binding pocket and shown as yellow sticks. The secondary structures within N-domain (Met1-Lys226) and C-
domain (Cys259-Thr454) are coloured cyan and green, respectively. The linker between C-domain and N-
domain from residues Thr227–Val258 is coloured gray. A detailed active site view of UGT74AC1-UDPG-1a.
The catalytic residues His18 and Asp111 are shown as magenta sticks. Yellow dotted lines indicate hydrogen
bonds. (B) The S 2 mechanism of UGT74AC1 catalysis. His18, with the aid of Asp111, specifically
N
deprotonates the 3-hydroxyl of mogrol and makes it a nucleophile to attack the sugar donor UDPG. (C) The
glycosylation of mogrol (1a) catalyzed by UGT74AC1 using UDPG as sugar donor. And specific activity of
UGT74AC1 wild type (WT) and mutants (H18A and D111A) in catalysis sites.
1
26
The N-terminal domain interacting with the acceptor consists of seven α-helices and seven
β-strands. Docking and MD simulation with 1a and complex enzyme-UDPG showed that 1a
127
ACS Paragon Plus Environment
6

Page 7 of 34
ACS Catalysis
1
2
3
1
1
28
29
could form a hydrogen bond with His18 at the acceptor pocket (Figure 1A). The dyad, His18-
Asp111, is located close to the sugar donor and the acceptor in SgUGT74AC1. Mutations of
His18 and Asp111 to alanine completely eliminated the activity on mogrol (Figure 1C). We
proposed that the residue His18 as the general base abstracts a proton from the 3-hydroxyl group
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
130
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
131
132
133
134
135
of 1a, and the generated nucleophile attacks the C1 carbon of glucose in UDPG, using S 2-like
N
mechanism.32,34,35 Site Asp111 stabilises the catalytic conformation and balances the charge
(Figure 1B). The inactivity of the His18 and Asp111 mutation indicated their important catalytic
role and confirmed that these two residues are irreplaceable.
1
36
2.2 Mutations for Improving Activity. To increase the catalytic efficiency of SgUGT74AC1
in the glycosylation of 1a, directed evolution and sequence/structure-based design approaches
were employed (Figure 2). The error-prone PCR method was executed initially to identify “hot
sites”. Mutation library containing approximately 5000 clones were selected based on a colour
change-based screening method,39 and five activity-improved mutants (L109I, R28H, F203I,
L48M and T79Y) were obtained (Figure 2A). The best hit in this round of screening was variant
T79Y which exhibited an eightfold enhancement in specific activity. These five residues are
located at the N-domain, and the distance between residue T79 and substrate 1a is within 4 Å.
Subsequently, iterative mutagenesis approaches were applied to test their synergistic effect.
Mutant T79Y served as the template to combine with four other mutants in the second-round
screening. Mutant M2 (L48M/T79Y) was chosen due to its 48-fold improvement in specific
activity compared with WT and served as the template for the sequential incorporation of the
remaining sites (R28H, L109I and F203I) (Table S2). The preferable mutant M4
(T79Y/L48M/R28H/L109I), which has a 379-fold higher catalytic efficiency (k /K value) to
137
1
1
38
39
140
141
142
143
144
145
146
147
1
1
48
49
cat
m
150
1a than WT, was discovered (Table 1). Although the three mutations (L48M, R28H and L109I)
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
51
52
are located far (12–21 Å) from 1a, their incorporation in M4 remarkable amplified the specific
activity relative to that of mutant T79Y, indicating that the residues without forming a direct
interaction with the substrate also influence the enzymatic activity of UGTs. F203I and L109I
mutations based on M3 (T79Y/L48M/R28H) led to improved activity, respectively, but their
combination in M5-1 exhibited a negative effect (Figure 2A).
153
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
54
55
A
M1 : T79Y
M1-1: L48M
M1-2: F203I
M1-3: L109I
M1-4: R28H
M2 : T79Y/ R28H
M2-1: T79Y/ L48M
M2-2: T79Y/ L109I
M2-3: T79Y/ F203I
M3 : T79Y/ R28H / L48M
M3-1: T79Y/ R28H / L109I
M3-2: T79Y/ R28H / F203I
M4
200
WT
M4-1
190
M3
96
M5-1
96
Random
mutagenesis
M2
48
M2-3
32
M3-1
M2-1
4
8
4
0
M2-2
24
M3-2
24
M1
M1-1
5
M1-2
M1-4 2
2
8
M4 : T79Y/ R28H /
L48M / L109I
M4-1: T79Y/ R28H /
L48M / F203I
M1-3
M4
2
M 5 : T 7 9 Y / R 2 8 H
/L48M / L109I/ F203I
WT
1
T79Y/ R28H /
L48M / L109I
4
B
i
ii
4
3
2
1
0
2
0
Alanine scanning
mutagenesis near
binding pocket
M4 S15A S15V S15K S15Y S15H S15G S15P S15D
M5
S15A/T79Y/
R28H / L48M
/L109I
1
C
P
F
P
*
D
4
3
2
1
0
Sequence
F
M
0.9
MΔ
D
Δ
consensus analysis
M
D
E
T
L
K
Q
H
I
0
.8
DΔ
E
T
L
0.7
0.6
0.5
0.4
T
*
*
LΔ
L
K
Q
H
L
*
M7
Q
H
Δ
Δ
H
I
*
H47R/M76L/
S15A/T79Y/
R28H / L48M
I
V
N
S
I
I
0
.3
V
N
S
Δ
V
G
V
*
0.2
0.1
0
SΔ
SΔ
S
S
S
/
L109I
*
R
G
A
R
G
R
R
Δ
G
*
*
*
A
A
A
A
A
A
P
5
A
N
R
K
L
L
A
87
S26
44
4
7
52
76
127
1
5
26
44 47
49
Position
52
76 87 109 127
M
H
49
V
G
M
M
D
Q
1
56
1
1
57
58
Figure 2. Evolution workflow of UGT74AC1. (A) Five mutants (L109I, R28H, F203I, L48M and T79Y) are
screened from approximately 5,000 clones, followed by combination of five sites in different ways. The
ACS Paragon Plus Environment
8

Page 9 of 34
ACS Catalysis
1
2
3
1
1
1
1
1
1
1
1
1
1
1
59
60
61
62
63
64
65
66
67
68
69
numbers below the horizontal lines show the relative activity of randomisation mutants of UGT74AC1 (WT)
and combined mutants compared with the WT. The best combination route is presented in the pink line,
leading to the best mutant M4 (T79Y/R28H/L48M/L109I, ~200-fold higher activity than WT) in the first
round. (B) Relative activity of alanine-scanning mutants of UGT74AC1-M4 around 4 Å to mogrol. Mutant M5
(T79Y/R28H/L48M/L109I/S15A, ~3.8-fold higher than M4) is obtained from alanine-scanning mutagenesis
near the binding pocket. Error bars represent standard deviation from three repeats. (C) and (D) Active-based
sequence conservation analysis of N-terminal of 13 UGTs that show activity to mogrol or mogrosides. (C) The
frequency distribution of amino acid occurrence for mutated targets. * and Δ stands for the mutants and amino
acid in WT at each point, respectively. (D) Relative activities of the selected mutants of UGT74AC1-M5
(UGT74AC1-T79Y/L48M/R28H/L109I/S15A). Error bars represent standard deviation from three repeats.
The relative activities of mutants were determined on crude extracts.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
170
Inspired by the beneficial mutation T79Y, we focused on the mutations near the catalytic
centre. Nineteen residues (F13, P14, S15, G17, D69, R70, T73, T75, M76, T79, L80, T113,
Q134, L138, Y175, F177, I185, L188, L189) adjacently located (<4 Å) were selected for alanine
scanning mutagenesis with M4 and WT as the template (Figure S4), and the results showed that
additional S15A mutation resulted in 3.8-fold and 5.1-fold increased activity than M4 and WT,
respectively (Figure 2B-i and Table S3). Then, mutants of M4 having S15 replaced with seven
other amino acids, including valine, lysine, tyrosine, histidine, glycine, proline and aspartic acid
were investigated. The result (Figure 2B-ii) showed that S15Y, S15K and S15V displayed 1- to
2-fold increased activity than M4. Especially, mutant M5 (T79Y/L48M/R28H/L109I/S15A)
1
1
1
1
1
1
71
72
73
74
75
76
177
1
1
78
79
3
showed a 5.15-fold decrease in K value and a 5.57×10 -fold increase in k /K value, compared
m
cat
m
180
with WT (Table 1 and Figure S5, S6), suggesting that M5 showed higher affinity and
1
1
1
81
82
83
considerably improved catalytic efficiency to 1a than WT.
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
84
85
Table 1. Apparent kinetic parameters of UGT74AC1 and its mutants towards mogrol (1a) and
UDPG.
−
1
–1 –1
Folda
Substrate Enzyme
K (μM)
m
kcat (s )
k / K (mM ·s )
cat m
WT
M4
49.40 ± 5.9
70.20 ± 4.5
9.59 ± 0.5
15.10 ± 0.3
10.10 ± 0.3
890.20 ± 5
64.10 ± 3
1.02 × 10–4
0.05 ± 0.005
0.11 ± 0.03
0.55 ± 0.06
0.86 ± 0.04
1.48×10–4
2.06×10–3
1
3.79×102
0.78 ± 0.08
11.47 ± 0.5
36.42 ± 0.7
86.00 ± 6.6
1.67×10–4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Mogrol
UDPG
M5
M6
M7
WT
M4
M5
M6
M7
5.57×103
1.77×104
4.17×104
1
4.67×103
0.05 ± 0.004
0.10 ± 0.002
0.37 ± 0.01
0.38 ± 0.03
0.78 ± 0.05
1.51 ± 0.04
1.68 ± 0.02
2.56 ± 0.06
9.06×103
66.08 ± 6
219.50 ± 9
148.50 ± 5
1.01×104
1.53×104
1
1
1
86
87
88
Notes: WT: wild-type; M4: T79Y/L48M/R28H/L109I; M5: T79Y/L48M/R28H/L109I/S15A;
M6: T79Y/L48M/R28H/L109I/S15A/M76L; M7: T79Y/L48M/R28H/L109I/S15A/M76L/H47R.
a
Fold change over catalytic efficiency of WT.
1
89
90
Amino acids positively affecting the function of protein would survive preferentially during
natural selection.44,45 Consensus sequence design, which replaced non-consensus residues with
consensus ones at each positions after multiple sequence alignment (MSA), is a powerful
strategy to improve protein thermostability.46 In this work, we employed an activity-based
sequence conservative analysis (ASCA) approach to increase the catalytic efficiency of
UGT74AC1. Twelve plant UGT sequences (seven from Siraitia grosvenorii and five from
Arabidopsis thaliana) characterised by their glycosylation activity to 1a or its glycosides were
1
191
1
1
92
93
194
1
1
95
96
4
7
collected and aligned with UGT74AC1 by using MSA (Figure S1). Analysis of the degree of
conservation and accessibility of amino acids interacting with 1a in the N-terminal domain
helped to identify ten residues that had the highest frequency at individual positions. Besides
S15A and L109I as mentioned in mutant M5, eight other mutations were executed with M5 as
template (Figure 2C). The results in Figure 2D showed that the additional mutation of H47R and
M76L displayed 1.95- and 3.20-fold higher activities than M5, respectively. The k /K value of
197
1
1
2
98
99
00
201
cat
m
ACS Paragon Plus Environment
1
0

Page 11 of 34
ACS Catalysis
1
2
3
2
2
02
03
resulting mutant M6 (T79Y/L48M/R28H/L109I/S15A/M76L) was 3.17-fold higher than that of
M5. Variant M7 (T79Y/L48M/R28H/L109I/S15A/M76L/H47R) was disclosed to be the best
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
204
mutant with a 4.17×10 -fold increase in catalytic efficiency to 1a compared with WT (Table 1).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
2
2
2
05
06
07
08
09
The mutation sites were located at the N-terminal of the enzyme (Figure S7). The ASCA
approach gave four positive mutations (S15A, L109I, M76L and H47R), and their combination
in M7 led to a 96-fold increase in specific activity compared with M3 (Table S2). This result
showed that the ASCA method could be an effective and promising tool to engineer UGTs to
enhance the catalytic efficiency from small-sized mutation libraries.
2
10
2.3 The Molecular Basis of Enhanced glycosylation Activity. To gain an insight into the
source of high catalytic efficiency of the enzyme mutants to 1a, we obtained the crystal
structures of mutant M7 (PDB ID: 6L8W, 2.05 Å) and performed molecular docking and
unconstrained MD simulations (200 ns) of complexes WT-UDPG-1a and M7-UDPG-1a (Figure
211
2
2
12
13
4
8
214
S8). According to UGT mechanism, a catalytic conformation, which supports catalysis, occurs
when the distance between the 3-hydroxyl-O of 1a and the NE2 nitrogen of the catalytic residue
His18 is less than 3.6 Å and the angle of NE2/3-hydroxyl-H/3-hydroxyl-O is larger than 135°.
The last 100 ns MD trajectory with 10000 frames in total was utilized to calculate the emerging
2
2
2
2
2
2
15
16
17
18
19
20
4
9
frequency of catalytic conformations for each complex. The results showed that the frequency
value in complex M7-UDPG-1a was 74.93 %, approximately five times that observed in WT-
UDPG-1a (15.09 %) (Figure S9). This finding revealed that M7-UDPG-1a is more favourable
than WT-UDPG-1a for the formation of catalytically competent poses to produce corresponding
glycosides. Catalytic conformations of WT-UDPG-1a and M7-UDPG-1a are used to calculate
the binding free energy by using molecular mechanics generalised Born surface area
221
2
2
22
23
5
0
224
(MM/GBSA) method implemented in AMBER16. The calculated binding free energies for 1a
ACS Paragon Plus Environment
1
1

ACS Catalysis
Page 12 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
25
26
in complexes WT-UDPG-1a and M7-UDPG-1a were −41.3 and −49.9 kcal/mol, respectively.
The lower binding free energy of M7-UDPG-1a is in line with the experiment result that M7
have higher affinity than WT towards 1a.
227
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
A
M76
I185
T79
L80
Mogrol
L138
T113
F13
UPDG
H18 S15
B
L76
I185
Mogrol
L138
Y79
L80
T113
UPDG
F13
A15
H18
228
229
Figure 3. Catalytic conformations of WT-UDPG-1a complex (A) and M7-UDPG-1a complex (B) in MD
simulations. Mogrol and UDPG are shown as orange and blue sticks, respectively. Mutant residues are shown
as green sticks and amino acids around 1a are shown as blue sticks. Yellow dotted lines indicate hydrogen
bonds. Hydrophobic interactions are shown as magenta dashes.
2
30
231
32
2
233
The structural characteristics of catalytic conformation of complexes WT-UDPG-1a and
M7-UDPG-1a were then studied to explore the conformational changes. Their most
representative conformations were obtained by aligning and clustering the MD trajectory into
three clusters based on the backbone atoms of the protein. The representative catalytic
conformation of WT-UDPG-1a showed that residues His18 formed hydrogen bonds with
hydroxyl groups at C3 positions of 1a, respectively, and hydrophobic interactions were observed
2
2
34
35
236
2
2
37
38
ACS Paragon Plus Environment
1
2

Page 13 of 34
ACS Catalysis
1
2
3
2
2
39
40
among F13, L80, I185 and 1a (Figure 3A). However, in M7-UDPG-1a, more hydrophobic
interactions between enzyme and 1a were observed than that in WT (Figure 3B). In addition to
the hydrophobic interactions of F13, T113, L138 and I185 with 1a, the replacement of Ser15
with an alanine resulted in a new hydrophobic interaction which facilitated nucleophile attacks.
The benzene ring of T79Y in M7 formed new hydrophobic interactions with 1a and anchored the
tail of 1a. Replacing flexible Met76 with leucine also resulted in additional hydrophobic
interaction, which stabilized the hydrophobic tail of 1a. These hydrophobic interactions resulted
in stronger binding affinity than WT-UDPG-1a which was reflected in the significantly increased
specific activity (Table 1).
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
241
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
242
243
244
245
246
247
HOH
HO
HO
2
C
HOH
HO
2
C
HOOC
HO
HO
A
O
O
O
HO
HO
HO
HO
O UDP
O UDP
UDP-Gal
O UDP
UDP-GlcA
UDP-Glc
B 3
2
000
000
2582
2514
WT
M4
M5
M6
M7
1000
6
73
6
4
2
00
00
00
326
10
5
3
.6
1
.6
1
N.D.N.D.0.1
N.D.N.D.N.D.N.D.N.D.
0
UDP-Glc
UDP-Gal
UDP-GlcA
2
48
2
2
2
49
50
51
Figure 4. Determinants of sugar donor specificity of UGT74AC1. (A) Structures of UDP-sugar. (B)
Comparison of the glycosylation activity of UGT74AC1 WT and mutants for UDP-Glc, UDP-Gal, and UDP-
GlcA. N.D.:No detected activity.
2
2
2
2
52
53
54
55
2.4 Sugar Donor Scope of SgUGT74AC1 and its Variants. The kinetic parameters of
UGT74AC1 and its mutants towards UDPG were further measured. Mutant M4
3
(T79Y/L48M/R28H/L109I) displayed 14-fold lower K value and 4.67×10 -fold higher k /K
m
cat
m
value than WT. M5 had the similar K value but increased k /K value compared to M4. The
m
cat
m
ACS Paragon Plus Environment
1
3

ACS Catalysis
Page 14 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
56
57
k /K value of M7 was 1.69-fold higher than that of M5 to give an excellent catalytic efficiency
cat
m
4
2
to UDPG, which was 1.53×10 -fold higher than that of WT (Table 1). Next, we investigated the
sugar donor scopes for UGT74AC1 and its mutants employing two other UDP-sugars with 1a as
acceptor (Figure 4). WT just accepted UDPG as glycosyl donor. And this enzyme showed no
activity to UDP-galactose (UDP-Gal) and UDP-glucuronic acid (UDP-GlcA). Surprisingly, the
S15A mutation creatively endowed mutant M5 the catalytic ability to glycosylate 1a with UDP-
Gal as a sugar donor. Moreover, the catalytic activity of M7 with UDP-Gal sugar donor
increased 36-fold compared to M5. Structure analysis of complex WT-UDP-Gal-1a after MD
simulation showed that the catalytic residue His18 formed a hydrogen bond with 4/6–OH of
galactose residue in WT-UDP-Gal-1a, thereby impeding the catalytic conformation of 1a (Figure
5A, S10). However, the binding of galactose residue was stabilized in M7-UDP-Gal-1a and 1a
was observed in the stable catalytic conformation (Figure 5B). A salt bridge and a hydrogen
bond were introduced between H47R and galactose. A psi–psi interaction was found between
W330 and the uracil ring of UDP-Gal. The binding free energy of M7-UDP-Gal-1a was −103.9
kcal/mol, which is 69.6 kcal/mol lower than that of WT (−34.3 kcal/mol). These structural
changes led to stable enzyme-substrate binding to expand sugar donor promiscuity.
258
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
259
260
261
262
263
264
265
2
2
66
67
268
2
2
2
69
70
71
ACS Paragon Plus Environment
1
4

Page 15 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
A
W370
B
W370
D372
Q373
D372
Mogrol
H18
Q373
L279
Mogrol
UDP-Gal
W351
S278
UDP-Gal
H18
S353
Q333
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
G17
Q16
G17
N352
S278
C331
E356
L334
R47
H47
W330
Y247
Y247
C331
C
D
M76
L80
I185
I185
L188
L76
L189
L189
L188
T79
R70
R82
R82
W370
W370
T79
11a
11a
D69
D69
R70
6.5Å
2
.8Å
F13
S15
F13
H18 S15
UPDG
H18
UPDG
2
2
72
73
Figure 5. (A) and (B) Representative binding conformations of WT-UDP-Gal-1a and M7-UDP-Gal-1a in MD
simulations, respectively. Mutant residues are shown as green sticks and amino acids that interact with donor
UDP-Gal are shown as blue sticks. Yellow, magenta and red dotted lines indicate hydrogen bonds,
hydrophobic interactions and salt bridges, respectively. Psi-psi interaction is shown in blue line. The substrate
1a (upper structure) and UDPG (lower structure) are shown as blue and orange stick models, respectively. (C)
and (D) Representative conformations of WT-UDPG-11a and M7-UDPG-11a in MD simulations, respectively.
The substrate 11a (upper structure) and UDPG (lower structure) are shown as orange and blue stick models,
respectively.
274
75
276
2
2
2
2
2
77
78
79
80
2
81
2.5 Triterpenes Acceptor Scope of SgUGT74AC1 and Variants. To test the application
scope of UGT74AC1 and their mutants in the glycosylation of other structurally different
tetracyclic triterpenes, we further examined the enzymatic reactions between 2a-11a and UDPG.
Delightfully, although WT UGT74AC1 only had slight catalytic ability to 2a–10a, the
282
2
83
84
2
2
4
285
engineered mutants displayed 10 –10 -fold increase in their catalytic ability (Figure 6, Table S4
ACS Paragon Plus Environment
1
5

ACS Catalysis
Page 16 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
2
2
86
87
and Figure S14–S25). Especially, the highest improvement (1.3×10 -fold) in specific activity to
10a was observed under the catalysis of mutant M6. Compounds 1a, 3a, 4a, 8a and 9a proved to
be good substrates for mutant M7, and substrates 5a–7a were highly suitable for mutant M5. It
was noted that substrates bearing alkyl or alkene group at tail, e.g. 1a-4a, were easily
glycosylated by mutants in high efficiency, in contrast to triterpenes bearing a carboxyl group at
various positions of the side chain, e.g. 5a-11a. The hydroxyl and carbonyl of triterpene skeleton,
such as 3a/4a and 7a/8a, also influenced the catalytic ability probably due to the change in
hydrophobic interactions between the enzyme and substrates. Unexpectedly, mutants M6 and M7
displayed catalytic activity towards dehydrotrametenolic acid (11a) which cannot be accepted by
WT. We noted that UGT51 mutant with seven mutations (M7-1) achieved 610-fold higher
288
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
289
290
291
292
293
294
295
39
2
2
96
97
enzyme activity to 3a than that of the wild type. However, this mutant showed no activity to 1a.
It was noted that the mutants in our study exhibited not only higher multiples (>3000-fold) to 3a
(Figure 6) but also larger sugar donor and triterpene acceptor promiscuity compared with their
work.
298
2
99
3
3
3
3
3
00
01
02
03
04
To gain insights into the basis for enzyme mutants to catalyze the glycosylation of different
triterpenes, the catalytic conformation of 11a was also explored. The orientation of 11a can
sustain a relatively stable catalytic conformation in M7-UGT-11a with frequency of catalytic
conformation for 86.67%, while it can’t form catalytic conformation in WT-UDPG-11a (Figure
S11, S12). In WT-UDPG-11a, the long distance between catalytic residues His18 and C3–OH of
11a (6.5 Å) resulted in non-catalytic conformations and non-activity (Figure 5C). However, the
mutations of M76L and T79Y formed hydrophobic interaction with the hydrophobic tail of 11a.
Mutations M76L and T79Y anchored the substrate in the proper orientation, pushing the C3–OH
of 11a being close to the catalytic site of His18 (2.8 Å) and enabling the catalysis (Figure 5D).
305
3
3
06
07
308
ACS Paragon Plus Environment
1
6

Page 17 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
09
3
3
10
11
Figure 6. Specific and relative activity of UGT74AC1 (WT) and its mutants using different substrates (1a–11a)
and UDPG as acceptor and donor, separately. “*” indicate new compounds.
3
3
3
3
12
13
14
15
The glycosylated products were prepared on a preparative scale and their structures were
analysed by HPLC-MS and NMR (Figure S14-S70). The glycosyl moieties were β–linked to the
C3 hydroxyl group of triterpenes, suggesting that the engineered proteins conserved the same
high regioselectivity as WT enzyme. Five new triterpenes glycosides 5c–9c, which have not been
achieved by other UGTs previously, were acquired. Those novel triterpene saponins may have
application potential in pharmaceutical study.
316
3
3
17
18
3. Conclusions
319
20
In summary, we have successfully engineered a plant UGT, SgUGT74AC1, to attain several
2
4
3
mutants that exhibited dramatically increased catalytic abiltity (10 -10 folds) in triterpene
ACS Paragon Plus Environment
1
7

ACS Catalysis
Page 18 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
3
21
22
glycosylation by means of combining error-prone PCR, structure-based semi-rational design and
activity-based sequence conservative analysis approaches. Moreover, compared with WT
enzyme, the engineered mutants also displayed extended donor and acceptor promiscuity while
maintaining the regioselectivity for C3 hydroxyl group of triterpene substrates. Crystal structure
analysis and MD simulations provided a better understanding of the origins of the improved
catalytic activity and substrate promiscuity of these UGT mutants. It is the hydrophobic
interactions between enzyme and substrate that play a critical role in anchoring triterpenes at the
active centre. This research has also provided a model and some guidance for the engineering of
other plant UGTs. Given that these mutant enzymes exhibited different catalytic activities for
different triterpenes, our future work will be focused on the engineering of the enzymes to suit
specific triterpenes and further improvement of their catalytic activities and application scopes.
323
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
3
3
3
3
24
25
26
27
28
29
330
3
3
31
32
4. Experimental section
333
4.1 Materials and chemicals
3
3
3
3
3
34
35
36
37
38
Terpenoids (mogrol (1a), mogroside IE (1c), Bryodulcosigenin (2a), protopanaxadiol (3a),
protopanaxatriol (4a), Lucidenic acid LM (5a), Ganoderic Acid I (6a), Ganoderic Acid C2 (7a),
1
Ganoderic Acid G (8a), Lucidenic acid C (9a), 16α-Hydroxydehydrotrametenolic acid (10a) and
Dehydrotrametenolic acid (11a)) were purchased from Chengdu Biopurify Phytochemicals. And
all other reagents were purchased from Sigma–Aldrich.
3
39
4.2 Heterogonous expression and purification of enzymes
340
The coding regions of UGT74AC1 were cloned from the fruits of S. grosvenorii and inserted into
the pET-32a expression vector. The gene of macrolide-inactivating glycosyltransferase (OleD)
variant TDP-16, which has been applied in high throughput screening of UGT74AC1 mutants,
3
3
41
42
ACS Paragon Plus Environment
1
8

Page 19 of 34
ACS Catalysis
1
2
3
3
3
43
44
was synthesized by Genscript (nanjing) Co., Ltd and inserted into expression plasmid pET21.
The recombinant vectors pET32a-UGT74AC1 or pET21-TDP-16 were then transformed into E.
coli BL21 (DE3) for expression. The strains harboring appointed plasmid were grown in LB
medium containing 100 µg/mL of ampicillin at 37 ºC. When OD600 reached 0.6-0.8, the cells
were induced with 0.4 mM isopropyl-β-D-thiogalactopyranoside (IPTG) at 16 ºC, 200 rpm for
20 h. The cell pellets were harvested by centrifugation at 5000 g for 10 min at 4 °C, then,
suspended in lysis buffer containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl and 25 mM
imidazole. After sonication, the cell debris was centrifuged at 17000 g and 4 ºC. The supernatant
was applied to Ni-NTA agarose affinity column to pure the recombinant proteins.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
345
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
346
347
348
349
350
351
352
4.3 Construction of UGT74AC1 mutant libraries and library screening
3
3
3
53
54
55
GeneMorph II Random Mutagenesis Kit (Agilent Technologies) was used to introduce mutations
during error-prone PCR process. The pET32a-UGT74AC1 plasmid served as template. All
procedures were performed according to manufactures’ protocol. Approximately 5000 mutants
were examined. Site-saturation mutations were applied using primers containing NNK codon at
specified sites. All primers used are provided in Table S5. About 100 colonies at one site were
screened. Colonies were cultured in 96 deep-well plates with 300 µL LB medium (100 µg/mL of
ampicillin) in each well at 37 ºC overnight. After that, 10 µL culture of each cell was transferred
to a new plate with 1mL LB containing 100 µg/mL of ampicillin, and incubated at 37 ºC for 3 h.
when OD600 reached 0.6-0.8, the culture was applied with 0.4 mM IPTG at 16 ºC for 24 h. The
cell was harvested and suspended in 200 µL lysis buffer (20 mM Tris-HCl, 0.3 mg/ml lysozyme,
pH 8.0) and suffered freeze-thaw three times. Cells were removed and the supernatant obtained
was used for high-throughput assay. The reaction was in 0.1 mL Tris-HCl (20 mM, pH 8.0)
containing 0.25 mM mogrol, 1 mM UDP-Glucose, 0.5 mM 2-chloro-4-nitrophenyl β-D-
356
3
3
57
58
359
3
3
60
61
362
3
3
3
63
64
65
ACS Paragon Plus Environment
1
9

ACS Catalysis
Page 20 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
3
66
67
glucopyranoside, 1% Tween 80, 11 µM OleD variant TDP-16 and 50 µL supernatant. The
reaction was incubated at 30 ºC for 20 min and monitored at 410 nm.51
3
68
4.4 Specific activities of UGT74AC1 and its mutants
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
369
To detect the activity of UGT74AC1 and its mutants, the UGT activity assay was conducted in
3
3
70
71
300 µL reaction buffer, including 10 mM UDPG, 1 mM MgCl , 50 mM Tris-HCl ( pH 8.0) and
2
0.8 mg purified enzyme of WT or ~100 µg purified enzyme of mutants (depending on the
activity of mutants). The reaction was incubated at 40 ºC for 10 min (mutants) and 3h (WT) and
then stopped by adding same volume of Methanol. The reactants were centrifuged and filtered
through 0.22 µm filter and analyzed by HPLC or HPLC-MS. A single unit of enzymatic activity
was defined as the amount of enzyme that consuming 1 µmol of substrates acceptor per min.
372
3
3
73
74
375
376
4.5 Kinetic analysis of UGT74AC1 and its mutants
3
3
77
78
For the kinetic analysis of UGT74AC1 and its mutants towards mogrol, the reactions were
measured in 300 µL reaction buffer (10 mM UDPG, 1 mM MgCl , 50 mM Tris-HCl pH 8.0 and
2
379
0.8 mg purified enzyme of WT or 100 µg purified enzyme of mutants) at 40 ºC for 4 h (WT) or
30 min (mutants), with the concentration of mogrol varied from 0.02-0.4 mM and then stopped
by adding same volume of Methanol. For the kinetic analysis of UGT74AC1 and its mutants
towards UDPG, the reactions were measured in 300 µL reaction buffer (0.2 mM mogrol, 1 mM
380
381
382
383
384
385
MgCl , 50 mM Tris-HCl pH 8.0, 0.8 mg purified enzyme of WT; 100 µg purified enzyme of
2
mutants) at 40 ºC for 4 h (WT) or 30 min (mutants), with the concentration of UDPG varied from
0.02-0.4 mM and then stopped by adding same volume of Methanol. Reaction mixture was
analyzed by HPLC or HPLC-ESI-MS, using a reversed-phase column (Ultimate C18 column, 4.6
mm × 250 mm, 5 µm particle, Welch, Shanghai, China). ESI methods were as described
386
3
3
87
88
before.34,35 For HPLC analyses, solvent A (H O with 0.1% formic acid) and solvent B
2
ACS Paragon Plus Environment
2
0

Page 21 of 34
ACS Catalysis
1
2
3
3
3
89
90
(acetonitrile with 0.1% formic acid) was utilized as the mobile phase at flow rate of 1 mL/min
and appointed UV wavelength and gradient program are listed in Table S6.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
91
4.6 Preparation of glycosides
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
392
The reactions were conducted in 200 mL reaction buffer (50 mg acceptor, 200 mg UDPG, 1 mM
3
93
94
MgCl , 50 mM Tris-HCl pH 8.0 and 1L cell lysates of M7) at 40 ºC for 5 h and same volume of
2
3
Methanol was added. Then, the reaction mixture was evaporated to 5 mL. To identify the
structure of products, glycosides were purified by a preparative HPLC system using reverse-
395
3
96
97
phase Ultimate C18 column and redissolved in dimethylsulfoxide-d . And finally, purified
6
3
glycosylated products were subjected to AVANCE III 600MHz spectrometer and characterized
1
13
398
399
by H NMR, C NMR, COSY, TOCSY, HSQC and HMBC.
4.7 Crystallization and structure solution
4
4
00
01
All crystallization experiments were conducted at 25 ºC using the sitting-drop vapor-diffusion
method. In general, 1 μL UGT74AC1 containing solution (25 mM Tris-HCl, pH 7.5, 150 mM
NaCl; 30 mg/mL) was mixed with 1 μL of reservoir solution in 48-well Cryschem Plates, and
equilibrated against 100 μL of the reservoir solution. The optimized crystallization condition of
UGT was 20% PEG8000, 0.2 M magnesium acetate tetrahydrate, 0.1M sodium acodylate pH 6.5.
Within 3 to 4 days, the crystals reached dimensions suitable for X-ray diffraction. The mutant
UGT74AC1 crystals were grown in the same condition as wild-type enzyme. The crystals in
complex with UDPG were obtained by soaking with mother liquor containing 5 mM
corresponding molecules for 3-5 h before data collection. All of the X-ray diffraction data sets
were collected at beam line BL17U1, BL17B1, BL18U1 and BL19U1 of the Shanghai
Synchrotron Radiation Facility (SSRF). The crystals were mounted in a cryoloop and soaked
with cryoprotectant solution (25% PEG8000, 5% Glycerol, 0.2 M magnesium acetate
402
4
4
4
4
4
4
03
04
05
06
07
08
409
4
4
10
11
ACS Paragon Plus Environment
2
1

ACS Catalysis
Page 22 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
4
12
13
tetrahydrate, 0.1M sodium acodylate pH 6.5) prior to data collection at 100 K. The diffraction
5
2
images were processed by using HKL2000. The crystal structure of wild-type UGT74AC1 was
5
3
54
414
solved by molecular replacement (MR) method with MrBUMP from the CCP4 using the
structure of the UGT85H2 from Medicago truncatula (PDB ID code 2PQ6, 27 % sequence
identity with UGT74AC1) as a search model, most of the residues were built by using
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4
4
4
4
4
15
16
17
18
19
20
5
5
56
Buccaneer and ARP/wARP, the further manual model building and refinement was carried
out using Refmac557 and Coot.58 Prior to structural refinements, 5% randomly selected
reflections were set aside for calculating Rfree as a monitor.59 The mutant and complex structures
were determined by MR method with Phaser60 using the refined UGT74AC1 structure as a
search model. The 2F -F difference Fourier map showed clear electron densities for most amino
421
o
c
4
4
22
23
acid residues. Subsequent refinements by incorporating ligands and water molecules were
according to 1.0 σ map level. Data collection and refinement statistics are summarized in Table
S1. All figures were prepared by using the PyMOL program (http://pymol.sourceforge.net/).
424
425
4.8 Molecular docking and molecular dynamics simulation
4
4
26
27
In order to shed light on possible reasons for high catalytic efficiency or new activity of mutant
M7, UDPG from crystal structure of 2ACW was import into the UDPG binding site of UGTs by
superimposition. The Glide module in Schrödinger program was used to dock substrates into the
428
61
4
29
30
enzymes. The molecular dynamics (MD) simulation was carried out with Amber 16 molecular
4
dynamics package. Substrates parameters were generated using the antechamber module and
6
2
431
GAFF2 with AM1-BCC charges. The complex system was solvated in TIP3P with rectangular
4
4
4
32
33
34
water box with a buffer distance of 10 Å. Neutralization of the system was achieved by adding
+
Na counter-ions. A two-stage geometry optimization was performed first for the water and ions,
then the entire system, using steepest descent algorithm for the first 2000 steps followed by
ACS Paragon Plus Environment
2
2

Page 23 of 34
ACS Catalysis
1
2
3
4
4
35
36
conjugated gradient algorithm for another 1000 steps. Then, the system was slowly heated up to
300 K and well equilibrated. A 10 ns MD simulation was carried out with a distance constraint
(3.0 Å, 50 kcal/mol) between the coordinating N atom of catalytic residue His18 and the O atom
of the 1a or O atom of 11a, thereby simulating the induced-fit process of substrate binding.
Finally, 200 ns unconstrained production MD was performed at 300 K and 1 atm with 2 fs
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
437
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4
4
4
4
4
38
39
40
41
42
43
6
3
integration time step. The ff14SB force field was used to model the protein and the general
6
4
65
AMBER force field (GAFF) for substrates. The SHAKE algorithm was employed to fix all
6
6
bonds involving hydrogen atoms and the particle mesh Ewald (PME) method was applied to
treat long range electrostatic interactions. Substrate binding free energies were calculated using
molecular mechanics generalized Born surface area (MM-GBSA) method implemented in
444
48
4
4
45
46
AMBER16. Catalytic conformation was defined according to the reaction mechanism and the
emerging frequency of catalytic conformations was calculated in the last 100 ns of MD
6
7
447
simulations. Protein–ligand interaction profiler and Pymol was used to visualize models and
4
4
4
4
48
49
50
51
construct graphical illustrative figures.
ASSOCIATED CONTENT
SUPPORTING INFORMATION
This information is available free of charge on the ACS Publications website.
4
4
52
53
Supporting tables and figures showing details of specific activity of variant, crystallographic data
collection, SDS-PAGE of pure protein, HPLC chromatograms, NMR analysis and molecular
dynamics simulations.
454
4
4
55
56
AUTHOR INFORMATION
Corresponding Author
ACS Paragon Plus Environment
2
3

ACS Catalysis
Page 24 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
4
57
58
* E-mail: sun_yx@tib.cas.cn; liu_wd@tib.cas.cn; lin_jp@tib.cas.cn
Author Contributions
4
4
4
4
4
59
60
61
62
63
‡ J.L. and J.Y. contributed equally to this work. They performed the protein engineering
experiments, analyzed the data and prepared the manuscript. N.S. and W.L. sovled the crystal
structure of SgUGT74AC1 and its mutant. J. L., P. L., and C. L. performed the MD simulation
and analyszed the data. Y. Z., S.M., and Y.C. purified the products and analysized their structure.
Y.S and Y.M designed the experiment and revised the manuscript.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4
64
65
Notes
The authors declare no conflict of interest.
4
4
66
67
ABBREVIATIONS
UDP, Uridine diphosphat; UGT, Uridine diphosphate glycosyltransferases; MD, molecular
468
dynamics; WT, wild type; UDPG, UDP-glucose; PSPG, plant secondary product
glycosyltransferase; MSA, multiple sequence alignment; MM-GBSA, molecular mechanics
generalised Born surface area; UDP-Gal, UDP-galactose; UDP-GlcA, UDP-glucuronic acid;
HPLC-MS, High Performance Liquid Chromatography-Mass Spectrometry; NMR: Nuclear
Magnetic Resonance.
4
4
69
70
471
4
72
4
4
73
74
ACKNOWLEDGMENT
This work was supported by the National Key Research and Development Program of China (20
475
19YFA0905100), Natural Science Foundation of China (No. 31771909), Tianjin
biotechnology innovation capacity improvement action.
synthetic
4
4
76
77
REFERENCES
ACS Paragon Plus Environment
2
4

Page 25 of 34
ACS Catalysis
1
2
3
4
4
78
79
(1) Hsu, C. H.; Hung, S. C.; Wu, C. Y.; Wong, C. H. Toward Automated Oligosaccharide
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Synthesis. Angew. Chem., Int. Ed. 2011, 50, 11872-11923.
4
4
80
81
(2) Tiwari, P.; Sangwan, R. S.; Sangwan, N. S. Plant Secondary Metabolism Linked
Glycosyltransferases: an Update on Expanding Knowledge and Scopes. Biotechnol. Adv. 2016,
34, 714-739.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
482
4
4
4
4
83
84
85
86
(3) Makino, T.; Shimizu, R.; Kanemaru, M.; Suzuki, Y.; Moriwaki, M.; Mizukami, H.
Enzymatically Modified Isoquercitrin, α-oligoglucosyl Quercetin 3-O-glucoside, is Absorbed
More Easily than Other Quercetin Glycosides or Aglycone after Oral Administration in
rats. Biol. Pharm. Bull. 2009, 32, 2034-2040.
4
4
4
87
88
89
(4) Liang, C.; Zhang, Y.; Jia, Y.; Wang, W.; Li, Y.; Lu, S.; Jin, J. M.; Tang, S. Y. Engineering a
Carbohydrate-processing Transglycosidase into Glycosyltransferase for Natural Product
Glycodiversification. Sci. Rep. 2016, 6, 21051.
490
91
(5) Prakash, I.; Markosyan, A.; Bunders, C. Development of Next Generation Stevia Sweetener:
4
Rebaudioside M. Foods 2014, 3, 162-175.
492
(6) Dai, L.; Li, J.; Yang, J.; Men, Y.; Zeng, Y.; Cai, Y.; Sun, Y. Enzymatic Synthesis of Novel
Glycyrrhizic Acid Glucosides Using a Promiscuous Bacillus Glycosyltransferase. Catalysts
2018, 8, 615.
4
4
93
94
4
4
95
96
(7) Liang, D.; Liu, J.; Wu, H.; Wang, B.; Zhu, H.; Qiao, J. Glycosyltransferases: Mechanisms
and Applications in Natural Product Development. Chem. Soc. Rev. 2015, 44, 8350-8374.
4
4
4
97
98
99
(8) Cantarel, B. L.; Coutinho, P. M.; Rancurel, C.; Bernard, T.; Lombard, V.; Henrissat, B. The
Carbohydrate-Active Enzymes Data-Base (CAZy): an Expert Resource for Glycogenomics.
Nucleic Acids Res. 2009. 37, D233–D238.
ACS Paragon Plus Environment
2
5

ACS Catalysis
Page 26 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
5
00
01
(9) Yonekura-Sakakibara, K.; Hanada, K. An Evolutionary View of Functional Diversity in
Family 1 Glycosyltransferases. Plant J. 2011, 66, 182-193.
5
5
02
03
(10) McArthur, J. B.; Chen, X. Glycosyltransferase Engineering for Carbohydrate Synthesis.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Biochem. Soc. Trans. 2016, 44, 129-142.
5
5
5
04
05
06
(11) Li, Y.; Baldauf, S.; Lim, E.-K.; Bowles, D. J. Phylogenetic Analysis of the UDP-
Glycosyltransferase Multigene Family of Arabidopsis thaliana. J. Biol. Chem. 2001, 276, 4338-
4343.
507
(12) Achnine, L.; Huhman, D. V.; Farag, M. A.; Sumner, L. W.; Blount, J. W.; Dixon, R. A.
Genomics-Based Selection and Functional Characterization of Triterpene Glycosyltransferases
from the Model Legume Medicago truncatula. Plant J. 2005, 41, 875-887.
5
5
08
09
5
5
10
11
(13) Mochida, K.; Sakurai, T.; Seki, H.; Yoshida, T.; Takahagi, K.; Sawai, S.; Uchiyama, H.;
Muranaka, T.; Saito, K. Draft Genome Assembly and Annotation of Glycyrrhiza uralensis, a
Medicinal Legume. Plant J. 2017, 89, 181-194.
512
5
13
(14) Cui, L.; Yao, S.; Dai, X.; Yin, Q.; Liu, Y.; Jiang, X.; Wu, Y.; Qian, Y.; Pang, L.; Gao, L.;
Xia, T. Identification of UDP-Glycosyltransferases Involved in the Biosynthesis of Astringent
Taste Compounds in Tea (Camellia sinensis). J. Exp. Bot. 2016, 67, 2285-2297.
514
5
15
5
5
5
16
17
18
(15) Kang, K. B.; Jayakodi, M.; Lee, Y. S.; Park, H.-S.; Koo, H. J.; Choi, I. Y.; Kim, D. H.;
Chung, Y.J.; Ryu, B.; Dong, Y. L.; Sung, S. H.; Yang, T.-J. Identification of Candidate UDP-
Glycosyltransferases Involved in Protopanaxadiol-Type Ginsenoside Biosynthesis in Panax
ginseng. Sci. Rep. 2018, 8, 1-10.
519
5
5
20
21
(16) Itkin, M.; Davidovich-Rikanati, R.; Cohen, S.; Portnoy, V.; Doron-Faigenboim, A.; Oren, E.;
Freilich, S.; Tzuri, G.; Baranes, N.; Shen, S.; Petreikov, M.; Sertchook, R.; Ben-Dor,
ACS Paragon Plus Environment
2
6

Page 27 of 34
ACS Catalysis
1
2
3
5
5
22
23
S.; Gottlieb, H.; Hernandez, A.; Nelson, D.; Paris, H.; Tadmor, Y.; Burger, Y.; Lewinsohn, E.;
Katzir, N.; Schaffer, A. The Biosynthetic Pathway of the Nonsugar, High-Intensity Sweetener
Mogroside V from Siraitia grosvenorii. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, E7619-E7628.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
524
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
25
(17) Dai, L.; Liu, C.; Zhu, Y.; Zhang, J.; Men, Y.; Zeng, Y.; Sun, Y. Functional Characterization
of Cucurbitadienol Synthase and Triterpene Glycosyltransferase Involved in Biosynthesis of
Mogrosides from Siraitia grosvenorii. Plant Cell Physiol. 2015, 56, 1172-1182.
526
5
27
5
5
5
28
29
30
(18) Yin, Q.; Shen, G.; Di, S.; Fan, C.; Chang, Z.; Pang, Y. Genome-Wide Identification and
Functional Characterization of UDP-Glucosyltransferase Genes Involved in Flavonoid
Biosynthesis in Glycine max. Plant Cell Physiol. 2017, 58, 1558-1572.
531
532
533
(19) Xie, K.; Chen, R.; Li, J.; Wang, R.; Chen, D.; Dou, X.; Dai, J. Exploring the Catalytic
Promiscuity of a New Glycosyltransferase from Carthamus tinctorius. Org. Lett. 2014, 16, 4874-
4877.
534
(20) Rahimi, S.; Kim, J.; Mijakovic, I.; Jung, K.-H; Choi, G.; Kim, S.-C.; Kim, Y.-J.
Triterpenoid-Biosynthetic UDP-Glycosyltransferases from Plants. Biotechnol. Adv. 2019, 37,
107394.
5
35
36
5
5
5
37
38
(21) Fan, B.; Chen, T.; Zhang, S.; Wu, B.; He, B. Mining of Efficient Microbial UDP-
Glycosyltransferases by Motif Evolution Cross Plant Kingdom for Application in Biosynthesis
of Salidroside. Sci. Rep. 2017, 7, 1-9.
539
5
40
(22) Dai, L.; Liu, C.; Li, J.; Dong, C.; Yang, J.; Dai, Z.; Zhang, X.; Sun, Y. One-pot Synthesis
of Ginsenoside Rh2 and Bioactive Unnatural Ginsenoside by Coupling Promiscuous
Glycosyltransferase from Bacillus Subtilis 168 to Sucrose Synthase. J. Agric. Food Chem.
2018, 66, 2830-2837.
541
5
5
42
43
ACS Paragon Plus Environment
2
7

ACS Catalysis
Page 28 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
5
44
45
(23) Thompson, A. M. G.; Iancu, C. V.; Neet, K. E.; Dean, J. V.; Choe, J.-Y. Differences in
Salicylic Acid Glucose Conjugations by UGT74F1 and UGT74F2 from Arabidopsis thaliana.
Sci. Rep. 2017, 7, 46629.
546
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
47
(24) Brazier-Hicks, M.; Offen, W. A.; Gershater, M. C.; Revett, T. J.; Lim, E.-K.; Bowles, D.
J.; Davies, G. J.; Edwards, R. Characterization and Engineering of the Bifunctional N- and O-
Glucosyltransferase Involved in Xenobiotic Metabolism in Plants. Proc. Natl. Acad. Sci. U. S. A.
2007, 104, 20238-20243.
548
5
5
49
50
5
5
51
52
(25) Zong, G.; Fei, S.; Liu, X.; Li, J.; Gao, Y.; Yang, X.; Wang, X.; Shen, Y. Crystal Structures
of Rhamnosyltransferase UGT89C1 from Arabidopsis thaliana Reveal the Molecular Basis of
Sugar Donor Specificity for UDP-β-l-Rhamnose and Rhamnosylation Mechanism. Plant J. 2019,
99, 257-269.
553
5
54
5
5
5
55
56
57
(26) Wetterhorn, K. M.; Newmister, S. A.; Caniza, R. K.; Busman, M.; McCormick, S. P.;
Berthiller, F.; Adam, G.; Rayment, I. Crystal Structure of Os79 (Os04g0206600) from Oryza
sativa: A UDP-Glucosyltransferase Involved in the Detoxification of Deoxynivalenol.
Biochemistry 2016, 55, 6175-6186.
558
5
5
5
59
60
61
(27) Hiromoto, T.; Honjo, E.; Noda, N.; Tamada, T.; Kazuma, K.; Suzuki, M.; Blaber, M.;
Kuroki, R. Structural Basis for Acceptor-Substrate Recognition of UDP-Glucose: Anthocyanidin
3-O-Glucosyltransferase from Clitoria ternatea. Protein Sci. 2015, 24, 395-407.
5
5
5
62
63
64
(28) Hiromoto, T.; Honjo, E.; Tamada, T.; Noda, N.; Kazuma, K.; Suzuki, M.; Kuroki, R.
Crystal Structure of UDP-Glucose: Anthocyanidin 3-O-Glucosyltransferase from Clitoria
ternatea. J. Synchrotron Radiat. 2013, 20, 894-898.
ACS Paragon Plus Environment
2
8

Page 29 of 34
ACS Catalysis
1
2
3
5
5
65
66
(29) Offen, W.; Martinez-Fleites, C.; Yang, M.; Kiat-Lim, E.; Davis, B. G.; Tarling, C. A.;
Ford, C. M.; Bowles, D. J.; Davies, G. J. Structure of a Flavonoid Glucosyltransferase Reveals
the Basis for Plant Natural Product Modification. EMBO J. 2006, 25, 1396-1405.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
567
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
68
(30) Modolo, L. V.; Li, L.; Pan, H.; Blount, J. W.; Dixon, R. A.; Wang, X. Crystal Structures of
Glycosyltransferase UGT78G1 Reveal the Molecular Basis for Glycosylation and
Deglycosylation of (iso) Flavonoids. J. Mol. Biol. 2009, 392, 1292-1302.
569
5
70
5
5
5
71
72
73
(31) Li, L.; Modolo, L. V.; Escamilla-Trevino, L. L.; Achnine, L.; Dixon, R. A.; Wang, X.
Crystal Structure of Medicago truncatula UGT85H2–Insights into the Structural Basis of a
Multifunctional (iso) Flavonoid Glycosyltransferase. J. Mol. Biol. 2007, 370, 951-963.
5
5
5
74
75
76
(32) Shao, H.; He, X.; Achnine, L.; Blount, J. W.; Dixon, R. A.; Wang, X. Crystal Structures of
a Multifunctional Triterpene/Flavonoid Glycosyltransferase from Medicago truncatula. The
Plant Cell, 2005, 17, 3141-3154.
577
(33) Yang, T.; Zhang, J.; Ke, D.; Yang, W.; Tang, M.; Jiang, J.; Cheng, G.; Li, J.; Cheng, W.;
Wei, Y.; Li, Q.; Naismith, J. H.; Zhu, X. Hydrophobic Recognition Allows the
Glycosyltransferase UGT76G1 to Catalyze its Substrate in Two Orientations. Nat.
Commun. 2019, 10, 1-12.
5
5
5
78
79
80
5
81
(34) He, J.-B.; Zhao, P.; Hu, Z.-M.; Liu, S.; Kuang, Y.; Zhang, M.; Li, B.; Yun, C.-H.; Qiao,
X.; Ye, M. Molecular and Structural Characterization of a Promiscuous C-Glycosyltransferase
from Trollius chinensis. Angew. Chem. 2019, 131, 11637-11644.
582
83
5
584
(35) Hsu, T. M.; Welner, D. H.; Russ, Z. N.; Cervantes, B.; Prathuri, R. L.; Adams, P. D.;
Dueber, J. E. Employing a Biochemical Protecting Group for a Sustainable Indigo Dyeing
Strategy. Nat. Chem. Biol. 2018, 14, 256.
5
5
85
86
ACS Paragon Plus Environment
2
9