D-Tagatose manufacture through bio-oxidation of galactitol derived from waste xylose mother liquor

Article abstract of DOI:10.1039/c8gc00091c

In this study, the bio-oxidation of galactitol to the valuable d-tagatose is presented. This proposed strategy could start with the refined by-product of waste xylose mother liquor formed by chemical hydrogenation. Through a computationally guided enzyme screening approach, a robust polyol dehydrogenase (PdPDH) was rapidly identified from a massive number of candidates. When coupled with a water-forming NADH oxidase (StNOX) for environmentally benign cofactor regeneration, subsequent reaction optimization facilitated the complete transformation of 100 g L^-1 galactitol into the desired product within 15 h, with a yield of 91% in a 2 L scale preparative reaction. Compared to the current enzymatic isomerization system, our approach avoids low conversion, high operative temperatures and by-product formation, while enabling simplified product isolation.

Full text of DOI:10.1039/c8gc00091c

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: F. Sha, Y. Zheng,
J. Chen, K. Chen, F. Cao, M. Yan and P. Ouyang, Green Chem., 2018, DOI: 10.1039/C8GC00091C.
Volume 18 Number
7
7
April 2016 Pages 1821–2242
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1463-9262
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Page 1 of 10
PleaseGdr oe en no Ct ha de jmu si st t mr yargins
View Article Online
DOI: 10.1039/C8GC00091C
Journal Name
ARTICLE
D-tagatose manufacture through bio-oxidation of galactitol
derived from waste xylose mother liquor
Received 00th January 20xx,
Accepted 00th January 20xx
a
a,b
a
a
a,c
a
Feng Sha, Yucong Zheng, Jiao Chen, Kequan Chen, Fei Cao,* Ming Yan* and Pingkai
a
Ouyang
DOI: 10.1039/x0xx00000x
In this study, the bio-oxidation of galactitol to the valuable D-tagatose is presented. This proposed strategy could start with
www.rsc.org/
the refined by-product of waste xylose mother liquor formed by chemical hydrogenation. Through a computationally
guided enzyme screening approach, a robust polyol dehydrogenase (PdPDH) was rapidly identified from a massive number
of candidates. When coupled with a water-forming NADH oxidase (StNOX) for environmental benign cofactor regeneration,
-
1
subsequent reaction optimization facilitated the complete transformation of 100 g L galactitol into the desired product
within 15 h, with a yield 91% in a 2-L scale preparative reaction. Compared to current enzymatic isomerization system, our
approach avoids low conversion, high operative temperatures and by-product formation, while enabling a simplified
product
isolation.
none has eradicated such obstacles.
Introduction
The transformation of agro-industrial wastes into valuable
products has emerged as a “green” concept in sustainable
D-tagatose (2a), a nascent rare sugar sweetener, has a similar
1
4
1
industry. Xylose mother liquor (XML), generated during
xylitol manufacture, is known as a notorious waste but is also a
taste but only one-third of the calorific value of sucrose. It has
attracted increasing interests because it provides numerous
health benefits including anti-hyperglycemic activity, treatment of
1
5
resource enriched with multiple carbohydrates. Because of
,3 its complex composition (the saccharide part usually contains
2
diabetes, prominent prebiotic effects, and other bioactivities.
3
5-40% xylose, 10-15%
L-arabinose, 8-10% glucose, and 8-10%
Nowadays, commercial 2a is usually produced by isomerization
3
a
), direct separation of its components has proven to be very
6,17
of
D-galactose (3a) via L-arabinose isomerase (L-AI, EC
4
1
-7
tough and not cost-effective.
When used as crude
5.3.1.4).
However, due to the intrinsic thermodynamic
feedstock for fermentation, XML has been converted to
various compounds, but with low productivity and yield (Fig.
equilibrium of L-AI and the very similar properties between 2a
and 3a, limited conversion and costly purification still exist in
1
8-22
5
,6
S1 and Table S1, ESI†).
Recently, a significant advance has
2
been made owing to chemical hydrogenation (Fig. S2, ESI†).
current way.
Previous efforts have been devoted to
3
overcoming these challenges. Some examples include
screening thermostable L-AIs from hyperthermophiles thus to
In this route, polyols provide an attractive platform for XML
conversion to hydrogen, alcohols and hydrocarbons (Fig.
8
perform the isomerization reaction under higher temperture,
2
A).
4-26
1
Except galactitol (1a),
using protein engineering and enzyme immobilization
,10
techniques to enhance the performance of L-AIs,
9
adding
borate to cause an equilibrium shift through complex
1
1
formation with 2a
microbial degradation,
separation process using simulated moving bed (SMB)
,
removing residual 3a by selective
1
2
and developing an integrated
A
XML
H
2
sedimentation
1
3
chromatography. Nevertheless, to the best of our knowledge,
C5&C6 Polyols
Galactitol
Hydrogen
Alcohols
Hydrocarbons
?
B
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 2 of 10
View Article Online
DOI: 10.1039/C8GC00091C
ARTICLE
Journal Name
Figure 1. (A) XML refinery via polyols platform; (B) Synthesis of
The genome-mining approach is a basic and powerful tool
4
8-50
D-tagatose (2a) from galactitol (1a) by highly stereoselective for discovering new enzyme sources.
However, when
bio-oxidation.
facing the rapidly expanding sequence databases, in our case,
the hydrogenated product of 3a that has seldom reported
2
ΔGbind [kcal/mol]
7-31
1. Blast Searching
-2.00
-7.00 -12.00 -17.00 -22.00
applications,
is simply filtered from the polyol mixture due
PspPDH
VpPDH
-1
to its significantly lower solubility (about 33 g L , 25 °C) than
-1 32,33
other polyols (e.g. the solubility of xylitol is 640 g L ).
HtPDH
CjPDH
PdPDH
LaPDH
PssPDH
PsPDH
RmPDH
SspPDH
MsPDH
PpPDH
MlPDH
SaPDH
LrPDH
3
4
Based on the global xylitol market, it is estimated that at
least 10,000 metric tons of 1a could be refined annually as an
abundant, raw material, which inspires us to explore an
2. Homology Modelling
efficient functionalization of the by-product 1a
.
In recent years, enzymatic oxidation has rapidly developed
as a new, clean route in organic chemistry because it is highly
chemo-, regio-, and enantioselective, operates under ambient
conditions, does not require problematic solvents, and also has
theoretical yields of up to 100%, especially when utilizing
GfPDH
HpPDH
PadPDH
TtPDH
3. Substrate docking
DyfPDH
SinPDH
GlPDH
PhpPDH
DfPDH
CaPDH
KrPDH
SmPDH
PapPDH
GjPDH
SimPDH
SsPDH
+
nicotinamide-dependent oxidoreductases driven by NAD(P)
3
5-39
regeneration system.
Since oxidation of the hydroxyl at the
C2-position of 1a could directly lead to the sugar 2a, polyol
dehydrogenase (PDH, EC 1.1.1.-), which is usually involved in
polyols metabolism in organisms, was therefore emerged as an
ideal biocatalyst for the target transformation (Fig. 1B).
However, current studies of this area stalled at whole-cell
transformation by several microrganisms such as Arthrobacter
FnPDH
TcPDH
MhPDH
CbPDH
CspPDH
BspPDH
GvPDH
4. Molecular dynamics
MD binding energy
Selected PDHs
Template PspPDH
4
0
41
globiformis, Mycobacterium smegmati, and Enterobacter
4
2
the SDR family is composed of more than 60,000 non-
agglomerans, suffered from low availability, low substrate
loading, which is still far from “true” practical applications. To
this end, we screened out a new PDH with the desired
functional profile from a large number of candidates by a
computational guided strategy. Coupled with a water-forming
5
1
redundant sequences (data in 2010). Thus it is valuable to
find ways to narrow the screening pool and improve the hit
rate. Herein, we
Figure 2. The computationally guided data mining avenue. A
phylogenetic tree for the final 37 sequences as well as PspPDH
is depicted, along with their corresponding protein-ligand
interface energies. Five putative PDHs (red) were selected for
experimental characterization.
NADH oxidase from Streptococcus thermophiles,
a bi-
enzymatic oxidation system was constructed to directly
produce 2a from an environmental friendly route, with
satisfying productivity as well as simple downstream
purification.
present an integrated computationally guided method which
combined the bioinformatics and molecular modelling
systematically with functional constraints to rapidly obtain our
Results and discussion
Integrated computationally guided PDH screening
5
2
ideal biocatalysts (Fig. 2). First, the sequence of PspPDH with
a relatively high activity on 1a was input as a probe, and 125
representatives for this fold family were filtered from the
sequence dataset of 10,000 records generated by a concise
Microbial Protein BLAST search with the criteria of containing
As mentioned before, although several microorganisms could
conduct the reaction, amino acid sequences of the responsible
40-43
PDH are still unknown and tough to be cloned.
It is worthy
to note that the oxidation of 1a to 2a has been described
during the study of substrate spectrum of four enzymes which
belong to the short-chain dehydrogenases/reductases (SDR)
+
an active site motif (-YXXXK-) and a characteristic NAD(P) -
4
4
binding motif (-GXXXGXG-) (Table S2, ESI†). To broaden PDH
enzyme superfamily (RsPDH from Rhodobacter sphaeroides,
4
PspPDH from Pseudomonas sp., RcPDH from Rhodobacter
5
sequence diversity, 7 redundant sequences were removed by
5
3
4
6
47
setting the sequence identity cut-off of 80%. All potential
scaffolds with the desired function were assumed to be
structurally homologous to PspPDH. Since the crystal
structures for most of the selected proteins are not available,
five homology models of each were created from the Swiss-
Model server using templates identified by the target-
capsulatus and RlGDH from Rhizobium legumenosarum ).
Initial test was carried out to assess their catalytic
performance, all of four PDHs were cloned, and three
candidates were successfully expressed in Escherichia coli
BL21(DE3) (Fig. S4, ESI†). However, low activity or poor
thermostability was observed in all cases (Table 1). As the
known PDHs did not provide satisfactory characteristics, we
addressed the challenge of obtaining new robust PDHs.
5
4
template sequence alignment, and the best built model was
selected according to model quality estimation scores of
5
5
GMQE and QMEAN.
2
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 3 of 10
PleaseGdr oe en no Ct ha de jmu si st t mr yargins
View Article Online
DOI: 10.1039/C8GC00091C
Journal Name
ARTICLE
+
PdPDH was found to be strictly inherently NAD -dependent,
To enrich the potential PDHs likely to function on 1a, the
modelled substrate was then docked into the predicted active which might be cost-competitive for industrial applications
+
site of each energy minimized homology model in complex compared to other oxidoreductases with NADP dependence.
+ 56
with NAD(P) using AutoDock vina. According to a general The substrate scope of PdPDH was subsequently studied to
5
7
Tyr-based catalytic mechanism of the SDR family, the explore its versatility (Table 2 and Fig. S5, ESI†). PdPDH
sequence library was further refined to 37 by manual analysis displayed the highest and almost similar efficiency toward
m
of the predicted conformations of the C2-hydroxyl of 1a target polyol 1a and D-sorbitol (1b), with a kcat/K valuale of
-
1
-1
-1
-1
interacting
1.40s mM and 1.57 s mM , respectively. However, we
observed that PdPDH shows inhibition kinetics under high
Table 2. Kinetic Constants for identified substrates of PdPDH.
Table 1. Enzyme screening results for the oxidation of 1a.
Activity
Conv.
[%]
Dc
Half-life
[h]
Enzyme
-1 a
b
[α]
d
[U mg ]
K
m
k
cat
k
cat/K
m
-1
Substrate
-1
-1
PspPDH 3.81 ± 0.13
35
29
17
44
27
3
(D)
5.48 ± 0.54
16.87 ± 0.89
22.77 ± 0.81
30.63 ± 0.57
14.70 ± 0.58
[mM]
[s ]
[s mM ]
RsPDH
RcPDH
PdPDH
PsPDH
3.11 ± 0.16
3.24 ± 0.12
6.13 ± 0.21
2.90 ± 0.15
0.83 ± 0.11
(D)
1
1
a
b
6.01 ± 0.27
8.46 ± 0.12
1.40
(D)
-4.91 ± 0.12 (D)
-5.04 ± 0.06 (D)
+5.12 ± 0.06 (L)
5.95 ± 0.29
9.35 ± 0.14
4.07 ± 0.15
1.78± 0.04
0.18 ± 0.01
0.17 ± 0.01
0.074 ± 0.01
1.57
0.11
0.029
CaPDH
19.27 ± 0.61
a
b
Activity was measured at 30°C and pH 8.0 using crude cell lysate. Reaction 1c
38.72 ± 0.16
60.79 ± 0.87
44.60 ± 0.80
115.04 ± 3.61
84.35 ± 2.51
+
+
conditions: 1a (10 μmol), NAD(P) (15 μmol NAD for PspPDH, RsPDH, RcPDH,
+
PsPDH and PdPDH, 15 μmol NADP for CaPDH), crude cell lysate (50 μg) and Tris- 1d
HCl buffer (100 mM pH 8.0, 1 mL); the mixture was stirred in vessels for 1 hour at
c
D
3
0 °C. Conversion was determined by HPLC. optical rotation ([α] ) was
-3
1
1
1
e
f
4.1×10
determined by polarimetry and absolute configuration was assigned by
D
47
comparison of [α] with those reported in the literature. The half-life was
-3
1.5×10
estimated as the time in which the initial activities of the purified enzymes (0.5
-
1
mg mL ) incubated in Tris-HCl buffer (100 mM pH 8.0) at 40 °C were reduced by
-4
g
8.9×10
5
0%.
with the Ser-Tyr-Lys catalytic triad and cofactor in the ternary concertation of 1b (k
=315 mM, Fig. S5-b, ESI†). PdPDH also
-iditol (1c) and xylitol (1d),
-arabitol, -mannitol, and
1g). These results indicate that the configuration of
i
structure. These sequences shared 25-63% amino acid exhibited definite activity toward
L
5
8
identities and a TMalign score of >0.8, indicating that they but weak activity in oxidizing
shared approximately the same folds as PspPDH (Table S3, talitol (1e
L
D
D-
-
ESI†). We assumed that they could catalyze 1a selective the hydroxy group at C-3 is vital for substrate identification.
oxidation but with differences in affinity. Further, the The enzyme has a higher preference for the (R)-configuration
Molecular Mechanics Poisson-Boltzmann Surface Area method of C-2. However, shortening the carbon chain or changing the
from all-atom molecular dynamic simulations (MD) was configuration of the hydroxy group at C-4 was found to result
followed to estimate the binding free energy (ΔGbind) of each in lower activity. This may be due to the positions of the
5
9
ternary structure. As shown in Fig. 2, these enriched and hydroxyl groups at C-4 and C-5 providing a binding force with
diverse sequences as well as PspPDH are depicted in a the polar residues (Gln140, Gly146 and Gly183) in the active
5
6
phylogenetic tree, along with their corresponding protein- pocket (Fig. S6, ESI†). Thus, some pentitols or tetritols, such
ligand interface energies. Within different regions of identity, as -arabitol, ribitol, erythritol and -threitol, could not be
five oxidoreductases with the lowest energies were finally oxidized by PdPDH.
D
L
chosen for experimental characterization.
As expected, three of them (PdPDH, PsPDH and CaPDH) Application of NADH oxidase for NAD regeneration
gave positive results, displaying activities of 6.13, 2.90 and For bio-oxidation catalyzed by oxidoreductases, a crucial
+
-
1
0
.83 U mg crude protein in cell-free extracts, respectively bottleneck is the continuous supply of the nicotinamide
+
Table 1). In a 10 mM-scale reaction employing equal amounts cofactor (NAD(P) ), which acts as the electron acceptor.
(
of crude protein from cell lysate for the conversion of 1a at 30 Moreover, the presence of excess cofactor is required to drive
3
5,60-62
°
C, PdPDH gave the highest conversion of 44% in 1 h, as well as the reaction equilibrium towards the desired product.
As
+
-configuration. Compared to the mentioned before, PdPDH is a strictly NAD -dependent
affording a product in the
D
+
other candidates, PdPDH offers best balance of activity and biocatalyst. Some familar enzyme-mediated NAD recycling
thermostability, with a half-life of 30.63 h at 40 °C, which is systems such as acetaldehyde/alcohol dehydrogenase,
almost 5.5-fold longer than that of PspPDH, and this property pyruvate/lactate dehydrogenase, or 2-oxoglutarate/glutamate
of PdPDH makes it a good candidate for further detail study.
dehydrogenase always lead to the formation of stoichiometric
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 4 of 10
View Article Online
DOI: 10.1039/C8GC00091C
ARTICLE
Journal Name
quantities of by-products, thus equally complicating the ions, and even slight activation was observed upon the
63
. Alternatively, NADH oxidase is more addition of 1 mM EDTA (Table S5, ESI†).Therefore, as a
isolation of product 2a
appropriate to couple with PdPDH due to it catalyzes the compromise, a mild pH of 8.0 and 30 °C were chosen as metal-
+
irreversible oxidation of NADH to NAD with simultaneous free reaction conditions for all subsequent experiments.
6
4
+ 69
As NAD is easily disrupted by alkali, we tested the
reduction of oxygen to either water or hydrogen peroxide.
+
Most species of the genus Streptococcus represent a stability of NAD (0.2 mM) in 100 mM Tris-HCl buffer (pH 8.0)
promising source of NADH oxidase, as an important part of for 24 h before main reaction parameters optimization, and
6
5-67
+
Herein, a new only a slight loss of 10% NAD was found under mild condition
protection mechanism against O
2
toxicity.
+
NADH oxidase (StNOX), identified from Streptococcus (Fig. S12, ESI†). Thus, decay of NAD is unconsidered over the
thermophilus, was successfully expressed in E. coli BL21(DE3) time scale of our experiments. In order to maximize the
+
(Fig. S4, ESI†). StNOX allowed rapid oxidation of NADH to NAD
performance of the bio-oxidation system, the PdPDH/StNOX
with the sole formation of innocuous water (Fig. S7, ESI†), ratio, catalyst dosage, oxidants, whole-cell biocatalysis, and
+
unlike other H
2
O
2
-forming NADH oxidases that require NAD concentration were
, which Table 3. Optimization of the enzyme-coupled reaction conditions.
a
application of catalase for in-situ elimination of H
2 2
O
6
could inactivate the enzymes. This protein shares high
4
+
b
6
5
sequence identity with known SmNOX (76.79%), SpyoNOX
PdPDH
U]
StNOX
[U]
NAD
Conv. [%]
Entry
[
[mM]
6
6
67
2 h
46
6 h
58
(75.54%),
and SpneNOX (71.49%)
from genus
c
1
60
60
0.2
0.2
0.2
0.2
0.2
0.2
0
Streptococcus. Alignment studies of the primary structure have
demonstrated that StNOX contains a pivotal catalytic residue
Cys44, an NADH-binding domain (-GXGXXG-), and two FAD-
binding domains (-GXXXXG-, and -TSXXDXXAXGD-) (Fig. S8,
ESI†).
c
2
60
120
240
480
480
480
480
480
55
64
75
93
94
75
95
86
86
93
98
99
83
99
c
3
60
c
4
120
120
120
120
120
c,e
5
Since StNOX is a typical flavoprotein oxidase (Fig. S9, ESI†),
we pre-incubated purified enzymes with exogenous flavin
cofactor for 10 min before activity assay. The maximum
activity with a 50% increase was achieved at an externally
added FAD of 10 μM, whereas continuously increasing FAD
d,e
6
d,e
7
d,e
8
0.1
a
+
Reaction conditions: 1a (1.5 mmol), NAD (0-2 μmol), PdPDH (60-120 U), StNOX
concentrations gradually decreased enzyme activity due to the (60-480 U), and Tris-HCl buffer (100 mM pH 8.0, 10 mL); the mixture was stirred
b
inhibition by FAD (Fig. S10, ESI†). This dependence on FAD in an open Erlenmeyer flask (50 mL) at 1000 rpm and 30 °C. Determined by HPLC
c
addition as well as release of FAD after holoprotein treatment analysis. purified enzymes, PdPDH (15.12 U/mg protein) and StNOX (18.10 U/mg
d
with hot water confirm that the flavin cofactor is not protein). Lyophilized cells of E. coli BL21 harboring PdPDH (5.40 U/mgCDW) and
6
8
e
covalently bound to the native enzyme as other oxidases. In StNOX (12.30 U/mgCDW), respectively. Under an oxygen atmosphere (O
2
-balloon).
the absence of external FAD, an acceptable enzyme activity of
-1
1
8.0 U mg protein was maintained at 30 °C and pH 8.0. The
kinetic parameters K and kcat for NADH were 0.14 mM and
6.83 s , respectively, where NADPH cannot serve as a then
m
-1
1
sequentially
optimized
in
a
10-mL
scale
substrate for this enzyme. Within the concentration range of biotransformation of 150 mM 1a. Different ratios of StNOX
.04-0.5 mM, StNOX was not inhibited by NADH (Fig. S5-h, and PdPDH that affect the main reaction orientation were first
ESI†). Besides, we also tested StNOX activity in the presence of investigated in an air atmosphere. As shown in Table 3, at a
0
-1
other reaction compounds. No inhibition was found up to 150 fixed loading amount of PdPDH (6 U mL ), the reaction was
mM 1a or 500 mM 2a
.
notably boosted by increasing the ratio of StNOX (entry 1, 2).
However, further elevation of StNOX/PdPDH ratio to 3 only
increased the initial velocity, and the reaction convesion
stayed at 86% in 6 h (entry 3). Then we doubled the dosages of
StNOX and PdPDH without varying any other conditions, which
led to a higher transformation of 93% of the substrate 1a
Reaction optimization
In the first set of experiments, the physicochemical properties
of PdPDH and StNOX were assessed. Similar to most reported
3
5,60-62
oxidoreductases,
PdPDH was found to be preferentially
(
entry 4).
Unlike aboved other coupled bio-oxidation systems in
which the main reaction can be efficiently driven by increasing
oxidized under alkaline conditions (pH 8.0-11.5) with an
optimum pH of 10.5 (Fig. S11, ESI†). On the other hand, StNOX
showed excellent activity under weakly acidic to neutral
conditions (pH 6.0-8.0), while pH increased to >8.5 resulting in
dramatic inactivation. Both PdPDH and StNOX were relatively
stable under ambient temperature conditions (Fig. S11, ESI†).
However, StNOX was found to be more thermotolerant than
PdPDH when incubated at temperatures higher than 40 °C,
which may due to its thermo-microbial origin. Moreover, the
activity of neither PdPDH nor StNOX was dependent on metal
7
0,71
the concentration of soluble co-substrates,
in the NOX-
coupled system, the solubility of the co-substrate oxygen in
water was generally poor (about 0.25 mM, 25 °C, 0.1 Mpa).
Thus, to ensure a more effective regeneration system, we tried
to carry out the reaction under an oxygen atmosphere. As
expected, the reaction with an enhanced oxygen transfer rate
proceeded smoothly, giving 93% conversion within 2 h (entry
5
). In comparison with purified enzymes, whole-cell catalysts
4
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 5 of 10
PleaseGdr oe en no Ct ha de jmu si st t mr yargins
View Article Online
DOI: 10.1039/C8GC00091C
Journal Name
ARTICLE
offer some unique advantages, such as ease of preparation Finally, to confirm the feasibility of this process in practical
and extended lifetimes of enzymes, and can further simplify applications, a preparative-scale reaction was carried out in a
7
reaction process. Meanwhile, whole-cell biocatalysis is much 7-
2
more likely to have undesirable metabolic by-products. L stirred bioreactor with an agitation speed of 500 rpm and air
-3
-1
Therefore, we incubated lyophilized E. coli BL21 cells with 1a flow rate of 0.36 m h (Fig. 3). By employing 9.7 g lyophilized
+
and 2a, respectively, over 24 h under the reaction conditions cells (24 KU PdPDH and 96 KU StNOX) plus 0.13 g NAD in 2 L
of entry 4, and no background reaction was found. When water (with the pH maintained at 8.0), 200 g 1a were
lyophilized cells of E. coli BL21 (harboring PdPDH and StNOX, completely converted to 2a within 15 h without generating
respectively) were used instead of purified enzymes, complete any by-products. At the stage of substrate saturation, the
conversion was achieved in 6 h (entry 6). Moreover, since whole- reaction rate was controlled by kcat, that make contribution to
cell catalysts usually contain a certain amount of cofactor, the a rapid and steady increase in conversion. The average
+
+
-1 -
concentration of NAD could be decreased to 0.1 mM for NAD - volumetric productivity (space-time yield) was up to 160 g L d
1
+
with a total NAD turnover number (moles of 2a/ moles of
+
NAD ) of approximately 5,400. Subsequent work-up gave the
pure product with 91% yield.
1
8
6
4
00
0
5
4
3
2
1
00
00
00
00
00
0
Furthermore, we were concerned about the environmental
impact of the established bio-oxidation system compromising
PdPDH and StNOX. An E-factor analysis for wastes generated in
reaction was performed (Table 4). However, systematic
comparison of enzymatic isomerization strategy is not easy,
especially when data of product purification process is not
0
0
8
,9
given.
Notably, a newly developed strategy by taking
advantages of the stereospecific and irreversible C-C bond
formation in cascade reaction which comprising transaminase
20
0
(
TAtca) and transketolase (TKgst) was recently reported,
resulting to another irreversible synthetic route of 2a. L-serine
1h), pyruvate (1i) and -threose (1j) were transformed to 2a
with high conversion (>95%) of 1j, which is thought to be
0
5
10
Time [h]
15
20
25
(
D
,
regeneration in-situ (entries 7 and 8).
7
3
greener than enzymatic isomerization. This interesting route
has been studied well for E-factor calculation, and thus the E-
factor of the cascade route was chosen as a comparison. As
shown in Table 4, solvent (water) as the reaction medium
constituted the main contributor to the E-factor. Due to the
efficiency of the cascade reaction was still very low (Table 5),
led to a very high cost of reaction water for preparing the
same amount of 2a compared to our strategy. Secondly, 1h
and 1i used in the first step of the cascade also significantly
contributed to the total waste formed since the ammonia
Figure 3. Reaction curves of bio-oxidation of 1a to 2a on a 2-L
scale. The conversion (●) and dissolved 1a concentration (●)
were determined by HPLC. The residual 1a in the reaction
system (△) was calculated based on formed 2a.
Table 4. Estimation of the generated waste in the irreversible
reactions performed on preparative scale.
TAtca/TKgst
PdPDH/StNOX
coupled system [g]
This work
a
74
Contribution
cascade system [g]
7
M. Lorillière et al.
donor is usually excessive added in transaminase reaction.
2
Whereas, by using PdPDH/StNOX coupled system, problematic
excessive addition of substrate can be avoided.
Reaction water
Substrate
215000
5733
63
9900
Considering the efficiency of the biocatalytic process is also
vital for practical application. We sumarized the previously
reported methods, as shown in Table 5, the traditional
isomerization route from 3a had to be performed under
relatively high temperature conditions (> 60 °C) to obtain high
conversions, which create challenges for both enzymatic
stability and removal of unwanted brown impurities. As wild
-
Cofactor
0.65
By-product
1617
160
-
b
TAtca/TKgst
-
c
PdPDH/StNOX
-
60
SUM
222573
9961
a
strains can consume 2a in vivo,
a
tediously long
Calculations were performed for L-serine, pyruvate and
D-threose as substrates
b
biotransformation of 1a , as well as low substrate loading,
would significantly lower the productivity. Obviously, such
problems also can be circumvented in our two-oxidase system.
Moreover, preliminary results indicate that a similar strategy
could also be used for the preparation of other valuable
functional sugars (Table S6, ESI†).
with a yield of 52% based on the work of M. Lorillière et al.(2017). TAtca and TKgst
c
were employed as purified enzymes. PdPDH and StNOX were employed as
lyophilized whole cells.
Preparative D-tagatose transformation on 2-L scale
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 6 of 10
View Article Online
DOI: 10.1039/C8GC00091C
ARTICLE
Journal Name
Experimental
General
Raw material galactitol refined from waste xylose mother
liquor was provided by Futaste Co., Ttd with a purity of >99%.
D
-sorbitol (99%),
98%), -sorbose (98%), xylitol (99%) and
purchased from Sigma-Aldrich (Shanghai, China). D-tagatose
99%), -arabitol (99%), -arabitol (99%), erythritol (99%) and
ribitol (99%) were purchased from Aladdin (Shanghai, China).
-talitol (98%) was purchased from TCI (Shanghai, China). β-
D
-fructose (99%),
D
-mannitol (98%),
L-iditol
(
L
L
-threitol (99%) were
(
L
D
D
+
Nicotinamide adenine dinucleotide phosphate (NADP ) and β-
+
Nicotinamide adenine dinucleotide hydrate (NAD ) were
purchased from Roche (Basel, Switzerland). Unless specified
otherwise, all the other chemical reagents were obtained
Table 5. Enzymatic synthesis of naturally rare 2a using various biocatalysts.
Substrate
Time
[h]
Temperature
Conv.
[%]
S. T. Y.
Substrate
Biocatalyst
loading
Ref.
73
-
1 -1 e
[°C]
[g L d ]
-
1
[
g L ]
-
2
a
1
3
h, 1i, 1j
a
16/8.8/6
1.8
60
80
60
30
35
30
28
28
30
72
20
>95
68
8.0×10
TAtca/TKgst cascade system
b
AraA
1.5
8
c
2
Gali152
500
20
24
155
50
50
24
48
15
46
2.3×10
2.6
9
1
a
Arthrobacter globiformis
Mycobacterium smegmatis
Enterobacter agglomerans
85
40
12
60
3.4
41
20
92
8.7
42
Gluconobacter oxydans DSM 2343
Acetobacter sp. MIM 2000/5
20
16
3.2
43
-
3
2.0
0.4
>99
4.0×10
1.6×10
43
d
2
PdPDH/StNOX coupled system
100
This work
a
b
c
d
Conversion rate was calculated based on Ij. L-AI from Thermotoga neapolitana. Immobilized L-AI from Geobacillus stearothermophilus. Lyophilized cells of E. coli
e
BL21 harboring PdPDH and StNOX, respectively. Space time yield (STY) was calculated based on analytical conversion.
with bound cofactor. The best model was eventually selected
commercially as analytically graded and used without further based on the GMQE and QMEAN scores that higher scores
1
13
purification. H NMR and C NMR were measured on a Bruker indicate higher reliability of the built 3D model. For the best
Avance spectrometer (Fällanden, Switzerland). Optical rotation model without bound cofactor, the cofactor was manually
measurements were taken on an automatic polarimeter (P810, placed by alignment with the other four models complexed
+
75
with NAD(P) in PyMol. Lastly, each model used as the
receptor for next substrate docking was submitted to an
Hanon Instruments, Jinan, China) at 25 °C in a 10 cm long cell.
7
6
energy minimization protocol with the Amber 16 package.
AutoDock Vina v1.1.2. was then used to predict the optimal
Integrative enzyme screening approach
BLAST searching. To obtain the PDH sequences, a concise
Microbial Protein BLAST search was selected and then
performed using a PSI-BLAST (Position-Specific Iterated BLAST)
program with PspPDH as the probe. Algorithm parameter of
Max target sequences was set as 10,000, and others with
default settings. 125 representative sequences were extracted
from the 10,000 searching results, and redundant sequences
were removed with Jalview by setting the sequence identity
7
substrate-docked conformation. Structure of ligand 1a was
7
7
8
sketched in ChemDraw and minimized in Chem3D. Input
coordinate files were prepared using the Autodock Tools that
will add charges, merge non-polar hydrogens and assign
appropriate atom types to the ligand and receptor. During
docking simulation, the receptor was set as rigid and the grid
5
3
box with dimensions of 20×20×20 Å was placed into the active-
site pocket. Exhaustiveness of the search was set to the value
of 20 while all other parameters were kept as default. The
optimized docked conformations of substrate were visualized
in PyMol for post-docking analysis to enrich the screening for
cut-off of 80%.
Homology modelling and substrate docking. Homology
models for each protein filtered from BLAST searching were
5
4
created using the Swiss-Model server. Frist, for each target
sequence, five template proteins in PDB were selected from
the template searching results generated by the server using
the target-template sequence alignment. These selected
templates should observe the following criteria: have a high
coverage, sequence identity >30% and at least one crystallized
1a activity.
Protein-ligand binding free energy estimation. The Molecular
Mechanics Poisson-Boltzmann Surface Area method from all-
atom molecular dynamic simulations (MD) was followed to
6
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 7 of 10
PleaseGdr oe en no Ct ha de jmu si st t mr yargins
View Article Online
DOI: 10.1039/C8GC00091C
Journal Name
ARTICLE
7
6
estimate the binding free energy of each ternary complex.
MD
Cells were harvested by centrifugation at 8,000×g for 10 min, 4
°C, washed twice with 10 mM Tris-HCl buffer (pH 8.0), and
simulations were performed by AMBER 16 package with stored at -80 °C until for further research (approximately 15 g
7
Amber protein.ff14SB force fields. The force field parameters L of cell dry weight).
9
-1
+
for substrate and NAD(P) was generated using the
Antechamber package and the partial charges were calculated
8
Enzyme purification
0
using the AM1-BCC method. Each simulation system was
solvated using SPC water box with a thickness of at least 10 Å,
Biocatalysts were purified to homogeneity by one-step nickel-
affinity chromatography. The frozen cell pellets were thawed
and resuspended in ice-cold binding buffer (500 mM NaCl, 10
mM imidazole, 5 mM 2-mercaptoethanol, 20 mM Tris-HCl, pH
+
-
and neutralized by addition of Na or Cl . These solvated
protein systems were first energy minimized for 10,000 steps
each to remove the close contacts and to relax the system.
After minimization, all the simulation systems were heated
with constant volume and temperature (NVT) for 10 ps from 0
K to 303 K and followed by production MD at 303 K for 20 ns.
According to the analysis of time dependence of the RMSD (Å)
of each simulation system (Fig. S3, ESI†), the MM-PBSA binding
free energies were calculated using 1,000 snap-shots retrieved
from the last 10 ns MD trajectories.
8.0) and disrupted by a high-pressure homogenizer (AH100B,
ATS, Ontario, Canada) at a pressure of 800 bars for 10 cycles.
Cell debris was removed by centrifugation for 30 min at
1
0,000×g and 4 °C. The supernatant was then purified on an
ÄKTAprime plus (GE Healthcare, Piscataway, USA) which
equip-ped with a HisTrap FF Crude column (column volume
(CV) = 5 mL, GE Healthcare, Piscataway, USA) that previously
equilibra-ted with binding buffer A. Unbound protein was
washed out after washing the column with 20 CVs of binding
buffer A, and His-tagged recombinant protein was eluted by
linear gradient elution with elution buffer (500 mM NaCl, 500
mM imidazole, 5 mM 2-mercaptoethanol, 20 mM Tris-HCl,
Phylogenetic tree of putative PDHs was inferred using the
8
Neighbor-Joining (NJ) method in MEGA and rooted at
1
PspPDH.
-1
B
iocatalysts expression
pH8.0) at a flow rate of 5 ml min . Fractions (each 5 ml)
The gene sequences of RsPDH, RlGDH, RcPDH, PspPDH, StNOX containing purified protein were pooled, desalted,
and selected candidates were codon optimized, synthesized, concentrated by Centriprep cells (10 kDa cut-off; Amicon, USA)
and subsequently cloned into EcoRI and HindIII restriction sites at 4 °C, with stock buffer (150 mM NaCl, 1mM dithiothreitol
of pET-28a(+) with an N-terminal 6
Suzhou, China), respectively. All the optimized gene sequences nitrogen, stored at -80 °C until for further research.
and their corresponding GenBank accession numbers were
The purity of enzymes was evaluated by 12.5% sodium
×
His tag in-frame (Genewiz, (DTT), 25 mM Tris-HCl, pH 8), and flash frozen in liquid
shown in supporting information Table S4. E. coli BL21(DE3) dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)
was chosen as a host for protein expression. The recombinant and stained with Coomassie Blue R-250. Protein concentration
-
1
cells were incubated in Luria-Bertani (LB) medium (10 g L
was determined with the BCA Protein Assay Kit (CWBIO,
-1
-1
Tryptone, 5 g L yeast extract, 10 g L Sodium chloride) which Beijing, China).
-1
contained 50 mg L kanamycin at 220 rpm and 37 °C until an
OD600 of 0.6 was reached, then induced by adding 0.2 mM
Lyophilized cells preparation
isopropyl-D-thiogalactoside (IPTG) with
cultivation at 16 °C.
a
further 18-h
Lyophilized cells were produced in 7-L bioreactor (see Biocatalysts
expression section). The frozen cell pellets were lyophilized under
vacuum at -50 °C using a Christ Alpha 2-4 LD plus freeze dryer
Large scale biocatalyst preparation was produced in a 7.0-L
stirred bioreactor (BioFlo 110, NBS, New Jersey, USA). The
recombinant cells were pre-cultured in 200 mL LB medium
(
Christ, Osterode am Harz, Germany), collected and stored at 4 °C
until use.
-1
with 50 mg L kanamycin at 220 rpm and 37 °C until the OD600
reached about 4, then inoculated into 2.8 L initial culture
-1
-1
medium (5 g L glycerol; 5 g L Tryptone; 5 g L yeast extract;
-1
Enzyme activity assays
-
1
-1
; 0.7 g L Na
-1
-1 For all the biocatalysts in this research were NAD(P)-
dependent oxidoreductases, the activity was determined
9
.0 g L Na
2
HPO
4
1
2
SO
4 2 4
; 3.4 g L KH PO ; 0.25 g L
-
MgSO ; 2.7 g L NH
4
4
Cl). The fermentation was carried out
+
based on NAD(P) reduction or NAD(P)H oxidation at 30 °C and
under 37 °C, with an aeration rate of 1 VVM, and pH 7.2 which
was controlled by ammonia water (2 M). Agitation speed was
changed between 200 and 500 rpm with dissolved oxygen
3
40 nm by a microplate reader (SpectraMax M5, Molecular
8
2,83
Devices, Santa Clara, USA), respectively.
Absorbance
readings were taken every 10 s for 3 min. A molar extinction
(DO) > 25%. After about 3-5 h, DO rebounded to 100% quickly,
-1
-1
coefficient of 6.22 mM cm was used for the calculation of
-1 enzyme activity. Briefly, the standard assay mixture for polyol
dehydrogenase was performed in a 200-μL system containing
indicating the consumption of carbon source in the initial
-
1
culture medium, and then feeding (250 g L glycerol; 60 g L
-1
Tryptone; 60 g L yeast extract) was started. When the OD600
reached about 8, the fermentation temperature was adjusted
to 20 °C and 1M IPTG stock was added to get final
concentration to 0.2 mM. The induction was continued for
about 14-18 h until the enzyme activity did not grow further.
1
00 mM Tris-HCl buffer (pH 8.0), 50 mM galactitol, 1 mM
+
NAD(P) and an appropriate amount of enzyme. For NAD(P)H
oxidase, the standard assay mixture (200 μL) comprised air-
saturated 100 mM Tris-HCl buffer (pH 8.0), 0.2 mM NAD(P)H,
and an appropriate amount of enzyme. The reaction was
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 8 of 10
View Article Online
DOI: 10.1039/C8GC00091C
ARTICLE
Journal Name
+
initiated by addition of enzyme. One unit of enzyme activity NAD was added into the reaction mixture, and samples were
was defined as the amount required for either reduction of 1 withdrawn at regular intervals for HPLC analysis. When oxygen
+
μmol NAD(P) or oxidation of NAD(P)H per minute under was used as oxidant, the flask containing reaction mixture
+
above standard conditions, respectively. Specific activity was except NAD was attached with the rubber septum and oxygen
expressed as units per milligram of protein. All measurements balloon. After flushing the flask with oxygen gas, the following
were performed in triplicate.
steps were the same as those described above for air as
oxidant.
Thermostability of the candidates
Each of purified candidates was diluted to 0.5 mg mL and Preparative-scale oxidation reaction
-1
incubated at 40 °C in 100 mM Tris-HCl buffer (pH 8.0). Samples Preparative-scale galactitol oxidation was carried out in a 7.0-L
were withdrawn at regular intervals and then their residual stirred bioreactor (BioFlo 110, NBS, New Jersey, USA)
oxidation activities were evaluated under standard conditions. containing 4.4 g lyophilized cells of PdPDH (24 KU), 5.3 g
The thermal inactivation half-life (t1/2) was defined as the time lyophilized cells of StNOX (96 KU), 200 g galactitol, and 0.13 g
+
at which the activity is reduced to 50% of the initial activity. NAD . The reaction was performed in water at agitation of 500
3
-1
The data were fitted to the eqn (1) and the first-order rpm and aeration rate of 0.36 m h with the pH maintained at
inactivation constants (k ) were determined by nonlinear 8.0 (2 M NaOH). The reaction temperature was controlled at
regression using OriginPro 2017.
d
30 °C, and samples were taken at regular time intervals and
·t) analyzed by HPLC. When conversion is complete, the reaction
mixture was centrifuged to remove the cells. After treatment
A
r
=
A
0
exp(-k
d
(1)
Where: A
0
= Initial enzymatic activity (U/mg of protein); A
r
=
with charcoal, the supernatant fluid was processed through
+
Dowex 50WX8 (Na ) and Dowex 1×2 (Cl ) to remove any
-
Residual enzymatic activity (U/mg of protein); and t= time (h).
The half-life was calculated from eqn (2).
+
impurities (e.g., NAD , and other ionic components including
t
1/2
=
ln
0.5/(-k
d
)
=
0.693/k
d
amino acids, peptides, and proteins), then lyophilized with an
isolated yield of 91% and fully characterized by NMR.
(2)
1
Characterization data for substrate galactitol. H-NMR (500 MHz,
Substrate scope of recombinant PdPDH and assay of enzyme DMSO-d ): δ = 4.37 (t, J = 5.6 Hz, 1H), 4.08 (d, J = 6.6 Hz, 1H), 4.04-
6
kinetic parameters
Substrate scope of recombinant PdPDH was explored by 3.36 (m, 2H). C NMR (126 MHz, DMSO-d ): δ =70.25 (s), 69.48 (s),
3.98 (m, 1H), 3.70 (q, J = 6.4 Hz, 1H), 3.45 (d, J = 6.3 Hz, 1H), 3.43-
1
3
6
1
testing various polyols (D-sorbitol, D-mannitol, L-iditol, xylitol, 63.26 (s). Characterization data for product
D-tagatose. H NMR (400
L-threitol, L-arabitol, D-arabitol, erythritol, ribitol and D-talitol) MHz, DMSO-d ): δ = 5.34 (s, 1H), 4.63 (d, J = 4.7 Hz, 1H), 4.52-4.44
6
at a fixed concentration of 50 mM and conducting enzyme (m, 2H), 4.40 (d, J = 4.4 Hz, 1H), 3.60-3.49 (m, 3H), 3.46 (dd, J = 11.0,
activity assays under standard conditions described above. The 7.0 Hz, 1H), 3.42-3.35 (m, 2H), 3.24 (dd, J = 11.0, 5.4 Hz, 1H).
1
3
PdPDH kinetic parameters (K
m
and kcat) for all identified
6
CNMR (150 MHz, DMSO-d ): δ = 97.88, 71.43, 70.65, 66.67, 64.54,
substrate were calculated from at least eight measurements 62.87.
by fitting initial velocity versus substrate concentration data to
the Michaelis-Menten equation. For
D-sorbitol, substrate
Determination of the biotransformation
Quantification of galactitol and -tagatose was performed
inhibition model was used to non-linear regression fitting.
Activity was measured at a varying substrate concentration of
D
on a HPLC device (Ultimate 3000, Dionex, USA) equipped with
the Aminex HPX-87H column (Bio-Rad, CA) and a RI detection
galactitol (2.5-100 mM),
00 mM), xylitol (5-200 mM),
mannitol (5-450 mM), and -talitol (5-450 mM), respectively.
D
-sorbitol (2.5-100 mM),
L-iditol (5-
2
L
-arabitol (5-200 mM),
D-
2 4
system (Shodex, Japan). 5 mM H SO was used as the mobile
D
-1
phase with a flow rate of 0.6 mL min . The column
temperature was maintained at 65 °C, and retention time of
In the case of NADH oxidase, the kinetic parameters were
obtained by varying NADH concentrations from 0.04 to 0.5
mM without exogenous FAD.
galactitol and
D
-tagatose were 10.32 and 9.63 min,
-tagatose in the
respectively. The content of galactitol and
D
samples was obtained directly using calibration curves.
Considering substrate saturation in the preparative-scale
reaction system, the eqn (3) was used to calculate the
conversion of the substrate.
Optimization of the PdPDH/StNOX coupled system
To fully maximize the catalytic properties of coupled system
towards galactitol, the reaction system was optimized in terms
of PdPDH/StNOX ratio, catalyst dosage, oxidants, whole-cell
+
Galactitol conversion = moles of
D-tagatose/
biocatalysis, and NAD addition. Typically, the reaction
moles of starting galactitol ×100% (3)
optimizations were carried out on a 10-mL scale in an open 50
mL Erlenmeyer flask. PdPDH and StNOX were firstly diluted in
1
00 mM Tris-HCl buffer (pH 8.0) with a designated ratio. The
mixture with galactitol added was then magnetically stirred at
0 °C at a constant speed of 1000 rpm to exclude the effect of In this study, we have demonstrated a novel green bio-
the stirring rate on the reaction. Time zero was recorded when oxidation system for
-tagatose production from the XML-
Conclusions
3
D
8
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 10
PleaseGdr oe en no Ct ha de jmu si st t mr yargins
View Article Online
DOI: 10.1039/C8GC00091C
Journal Name
ARTICLE
derived raw material galactitol.
A
combination of 16 Y. Xie, C. Y. Chin, D. S. C. Phelps, C. Lee, K. B. Lee, S. Mun and
N. L. Wang, Ind. Eng. Chem. Res., 2005, 44, 9904-9920.
bioinformatics and molecular modelling enabled the rational
identification of a highly stereo-selective PdPDH from a
massive sequence database. In particular, by coupling it with a
water-forming StNOX, the complete conversion of galactitol
1
1
7 H. Lei, Z. Bao, H. Xing, Y. Yang, Q. Ren, M. Zhao and H.
Huang, J. Chem. Eng. Data, 2009, 55, 735-738.
8 H. Wang, L. Li, L. Zhang, J. An, H. Cheng and Z. Deng, Microb.
Cell Fact., 2016, 15, 82.
with high substrate loading led to a considerably simplified 19 T. Zhou and J. Zhong, Bioresour. Technol., 2015, 175, 160-
1
66;
downstream process and minimum waste formation, therefore
suggesting its utility for practical production on a large scale.
Moreover, our work offers a good example of how to
significantly improve the overall economic feasibility of the
2
2
0 X. Liu, L. Lin, X. Xu, H. Zhang, L. Wu, P. Zhu, S. Li and H. Xu,
Carbohydr. Polym., 2018, 181, 412-418.
1 J. Guo, J. Li, Y. Chen, X. Guo and D. Xiao, Appl. Biochem.
Biotechnol., 2016, 180, 717-727.
current polyols platform where developments may provide 22 B. Tang, H. Xu, Z. Xu, C. Xu, Z. Xu, P. Lei, Y. Qiu, J. Liang and X.
Feng, Bioresour. Technol., 2015, 181, 351-354.
more attractive solutions, ultimately, to realize full and
2
2
3 CN Pat., 201010278356.3, 2010.
complete utilization of XML in the future.
4 J. M. Robinson, C. E. Burgess, M. A. Bently, C. D. Brasher, B.
O. Horne, D. M. Lillard, J. M. Macias, L. D. Marrufo, H. D.
Mandal, S. C. Mills, K. D. O’Hara, J. T. Pon, A. F. Raigoza, E. H.
Conflicts of interest
Sanchez, J. S. Villarreal, Q. Xiang, ACS Symp. Ser., 2006, 921
04-320.
5 G. W. Huber, J. W. Shabaker, S. T. Evans and J. A. Dumesic,
Appl. Catal., B, 2006, 62, 226-235.
6 E. L. Kunkes, D. A. Simonetti, R. M. West, J. C. Serrano-Ruiz,
C. A. Gärtner and J. A. Dumesic, Science, 2008, 322, 417-421.
7 A. Duclos, C. Fayet and J. Gelas, Synthesis, 1994, 10, 1087-
,
3
There are no conflicts to declare.
2
2
2
Acknowledgements
This work was supported by the Opening Foundation of State
1
090.
Key Laboratory of Materials-Oriented Chemical Engineering 28 I. Isaac, G. Aizel, I. Stasik, A. Wadouachi and D. Beaupere,
(
KL16-06), the Science and Technology Planning Project of
Synlett, 1998, 05, 475-476.
9 M. Kurszewska, E. Skorupowa, J. Madaj, A. Konitz, W.
2
3
Nanjing (201716035) and the Six Talent Peaks Project in
Jiangsu Province (SWYY-030). We are grateful to the High
Performance Computing Center of Nanjing Tech University for
supporting the computational resources. In addition, we would
Wojnowski and A. Wiśniewski, Carbohydr. Res., 2002, 337
261-1268.
0 B. S. Jeon, C. Moon, B. Kim, H. Kim, Y. Um and B. Sang,
,
1
Enzyme Microb. Technol., 2013, 53, 143-151.
like to thank Futaste Co., Ttd for supplying the raw material 31 A. Yamaguchi, N. Muramatsu, N. Mimura, M. Shirai and O.
Sato, Phys. Chem. Chem. Phys., 2017, 19, 2714-2722.
galactitol.
3
3
3
3
3
3
3
3
2 K. Schneider, G. Jakel, R. Hoffmann and F. Giffhorn,
Microbiology, 1995, 141, 1865-1873.
3 G. A. Evrendilek, in Varzakas T, Labropoulos A, Anestis S,
eds., 2012, ch.3, pp. 45-78.
4 L. V. Rao, J. K. Goli, J. Gentela and S. Koti, Bioresour. Technol.,
Notes and references
1
2
G. V. Levin, J. Med. Food, 2002,
T. W. Donner, J. F. Wilber and D. Ostrowski, Diabetes, Obes.
Metab., 1999, , 285-291.
Y. Lu, G. V. Levin and T. W. Donner, Diabetes, Obes. Metab.,
008, 10, 109-134.
5, 23-36.
2
016, 213, 299-310.
5 W. Kroutil, H. Mang, K. Edegger and K. Faber, Curr. Opin.
Chem. Biol., 2004, , 120-126.
1
8
3
6 F. Hollmann, I. W. Arends, K. Buehler, A. Schallmey and B.
Bühler, Green Chem., 2011, 13, 226-265;
2
4
5
6
P. Kim, Appl Microbiol Biot, 2004, 65, 243-249.
7 D. Romano, R. Villa and F. Molinari, ChemcatChem, 2012,
39-749.
8 M. G. López-Vidal and G. de Gonzalo, Adv. Synth. Catal.,
017, 359, 2026-2049.
4,
D. Oh, Appl. Microbiol. Biotechnol., 2007, 76, 1-8.
Z. Xu, S. Li, X. Feng, J. Liang and H. Xu, Appl. Microbiol.
Biotechnol., 2014, 98, 8869-8878.
7
2
7
8
9
1
1
1
J. Jayamuthunagai, P. Gautam, G. Srisowmeya and M.
Chakravarthy, Crit. Rev. Food Sci. Nutr., 2016, 10, 1126550.
B. Kim, Y. Lee, H. Lee, D. Lee, E. Choe and Y. Pyun, FEMS
Microbiol. Lett., 2002, 212, 121-126.
9 A. T. Martínez, F. J. Ruiz-Dueñas, S. Camarero, A. Serrano, D.
Linde, H. Lund, J. Vind, M. Tovborg, O. M. Herold-Majumdar,
M. Hofrichter, C. Liers, R. Ullrich, K. Scheibner, G. Sannia, A.
Piscitelli, C. Pezzella, M. E. Sener, S. Kılıç, W. van Berkel, V.
Guallar, M. F. Lucas, R. Zuhse, R. Ludwig, F. Hollmann, E.
Fernández-Fueyo, E. Record, C. B. Faulds, M. Tortajada, I.
Winckelmann, J. A. Rasmussen, M. Gelo-Pujic, A. Gutiérrez, J.
C. Del Río, J. Rencoret and M. Alcalde, Biotechnol. Adv.,
H. J. Kim, S. A. Ryu, P. Kim and D. K. Oh, Biotechnol. Prog.,
2
003, 19, 400-404.
0 Z. Zheng, W. Mei, M. Xia, Q. He and J. Ouyang, J. Agric. Food
Chem., 2017, 65, 4715-4721.
1 Y. Zhan, Z. Xu, S. Li, X. Liu, L. Xu, X. Feng and H. Xu, J. Agric.
Food Chem., 2014, 62, 2412-2417.
2 M. Liang, M. Chen, X. Liu, Y. Zhai, X. Liu, H. Zhang, M. Xiao
and P. Wang, Appl. Microbiol. Biotechnol., 2012, 93, 1469-
2
017,35, 815-831.
4
4
4
4
4
0 K. Izumori, T. Miyoshi, S. Tokuda and K. Yamabe, Appl.
Environ. Microbiol., 1984, 48, 1055-1057.
1 K. Izumori and K. Tsuzaki,
25-227.
J. Ferment. Technol. , 1988, 66,
1
474.
2
1
1
3 N. Wagner, A. Bosshart, J. Failmezger, M. Bechtold and S.
Panke, Angew. Chem. Int. Ed. Engl., 2015, 54, 4182-4186.
4 A. A. Koutinas, A. Vlysidis, D. Pleissner, N. Kopsahelis, I. L.
Garcia, I. K. Kookos, S. Papanikolaou, T. H. Kwan and C. S. K.
Lin, Chem. Soc. Rev., 2014, 43, 2587-2627.
2 S. Muniruzzaman, H. Tokunaga and K. Izumori, J. Ferment.
Bioeng., 1994, 78, 145-148.
3 M. Manzoni, M. Rollini and S. Bergomi, Process Biochem.,
2
001, 36, 971-977.
4 S. Schauder, K. Schneider and F. Giffhorn, Microbiology,
995, 141, 1857-1863.
1
5 H. Cheng, B. Wang, J. Lv, M. Jiang, S. Lin and Z. Deng, Microb.
Cell Fact., 2011, 10, 5.
1
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 10 of 10
View Article Online
DOI: 10.1039/C8GC00091C
ARTICLE
Journal Name
4
4
5 K. Schneider and F. Giffhorn, Enzyme Microb. Technol., 1991, 76 B. R. Miller III, T. D. McGee Jr, J. M. Swails, N. Homeyer, H.
1
3
, 332-337.
6 A. Sola-Carvajal, M. I. García-García, G. Sánchez-Carrón, F.
García-Carmona and Á. Sánchez-Ferrer, Biochimie, 2012, 94
407-2415.
Gohlke and A. E. Roitberg, J. Chem. Theory Comput., 2012, 8,
3314-3321.
,
77 O. Trott and A. J. Olson, J. Comput. Chem., 2010, 31, 455-
461.
2
4
4
4
5
5
5
7 S. S. Jagtap, R. Singh, Y. C. Kang, H. Zhao and J. Lee, Enzyme 78 ChemDraw Ultra, version 12.0.2, CambridgeSoft, 2010.
Microb. Technol., 2014, 58, 44-51.
8 Y. J. Zhang, W. X. Zhang, G. W. Zheng and J. H. Xu, Adv.
Synth. Catal., 2015, 357, 1697-1702.
79 D. A. Pearlman, D. A. Case, J. W. Caldwell, W. S. Ross, T. E.
Cheatham, S. DeBolt, D. Ferguson, G. Seibel and P. Kollman,
Comput. Phys. Commun., 1995, 91, 1-41.
9 H. Li, Z. J. Luan, G. W. Zheng and J. H. Xu, Adv. Synth. Catal., 80 J. Wang, W. Wang, P. A. Kollman and D. A. Case, J. Mol.
015, 357, 1692-1696. Graph. Model., 2006, 25, 247-260.
0 Y. P. Shang, Q. Chen, X. D. Kong, Y. J. Zhang, J. H. Xu and H. L. 81 K. Tamura, J. Dudley, M. Nei and S. Kumar, Mol. Biol. Evol.,
Yu, Adv. Synth. Catal., 2017, 359, 426-431. 2007, 24, 1596-1599.
1 Y. Kallberg, U. Oppermann and B. Persson, FEBS J., 2010, 82 B. R. Riebel, P. R. Gibbs, W. B. Wellborn and A. S. Bommarius,
77, 2375-2386. Adv. Synth. Catal., 2002, 344, 1156-1168.
2 W. S. Mak, S. Tran, R. Marcheschi, S. Bertolani, J. Thompson, 83 R. D. Woodyer, N. J. Wymer, F. M. Racine, S. N. Khan and B.
2
2
D. Baker, J. C. Liao and J. B. Siegel, Nat. commun., 2015,
0005.
6,
C. Saha, Appl. Environ. Microbiol., 2008, 74, 2967-2975.
1
5
5
3 A. M. Waterhouse, J. B. Procter, D. M. Martin, M. Clamp and
G. J. Barton, Bioinformatics, 2009, 25, 1189-1191.
4 M. Biasini, S. Bienert, A. Waterhouse, K. Arnold, G. Studer, T.
Schmidt, F. Kiefer, T. G. Cassarino, M. Bertoni and L. Bordoli,
Nucleic Acids Res., 2014, 42, W252-W258.
5
5
5
5 P. Benkert, S. C. Tosatto and D. Schomburg, Proteins: Struct.,
Funct., Bioinf., 2008, 71, 261-277.
6 O. Trott and A. J. Olson, J. Comput. Chem., 2010, 31, 455-
461.
7 A. Philippsen, T. Schirmer, M. A. Stein, F. Giffhorn and J.
Stetefeld, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2005,
61
, 374-379.
8 Y. Zhang and J. Skolnick, Nucleic Acids Res., 2005, 33, 2302-
309.
9 B. R. Miller III, T. D. McGee Jr, J. M. Swails, N. Homeyer, H.
Gohlke and A. E. Roitberg, J. Chem. Theory Comput., 2012,
314-3321.
5
5
2
8,
3
6
6
0 C. V. Voss, C. C. Gruber, K. Faber, T. Knaus, P. Macheroux and
W. Kroutil, J. Am. Chem. Soc., 2008, 130, 13969-13972.
1 P. Könst, H. Merkens, S. Kara, S. Kochius, A. Vogel, R. Zuhse,
D. Holtmann, I. W. Arends and F. Hollmann, Angew. Chem.
Int. Ed. Engl., 2012, 51, 9914-9917.
6
6
6
6
2 X. Jiao, Y. Zhang, Q. Chen, J. Pan and J. Xu, Catal. Sci.
Technol., 2016,
3 W. Hummel, M. Kuzu and B. Geueke, Org. Lett., 2003,
649-3650.
6, 7094-7100.
5,
3
4 G. Rehn, A. T. Pedersen and J. M. Woodley, J. Mol. Catal. B:
Enzym., 2016, 134, 331-339.
5 B. Petschacher, N. Staunig, M. Müller, M. Schürmann, D.
Mink, S. De Wildeman, K. Gruber and A. Glieder, Comput.
Struct. Biotechnol. J., 2014, 9, e201402005.
6
6
6 H. Gao, M. K. Tiwari, Y. C. Kang and J. Lee, Bioorg. Med.
Chem. Lett., 2012, 22, 1931-1935.
7 J. Yu, A. P. Bryant, A. Marra, M. A. Lonetto, K. A. Ingraham, A.
F. Chalker, D. J. Holmes, D. Holden, M. Rosenberg and D.
McDevitt, Microbiology, 2001, 147, 431-438.
6
6
8 Y. C. Zheng, J. H. Xu, H. Wang, G. Q. Lin, R. Hong and H. L. Yu,
Adv. Synth. Catal., 2017, 359, 1185-1193.
9 O. H. Lowry, J. V. Passonneau and M. K. Rock, J. biol. Chem.,
1
961, 236, 2756-2759.
0 W. Liu and P. Wang, Biotechnol. Adv.,2007, 25, 369-384.
1 Y. Zhang and H. Hess, ACS Catal. 2017, , 6018-6027.
2 J. Wachtmeister and D. Rother, Curr. Opin. Biotechnol., 2016,
, 169-177.
7
7
7
7
42
7
3 M. Lorillière, M. De Sousa, F. Bruna, E. Heuson, T. Gefflaut, V.
De Berardinis, T. Saravanan, D. Yi, W. Fessner and F.
Charmantray, Green Chem., 2017, 19, 425-435.
7
7
4 F. Guo and P. Berglund, Green Chem., 2017, 19, 333-360.
5 W. L. DeLano, The PyMOL Molecular Graphics System, v1.3,
Schrödinger LLC, 2002, http://www.pymol.org
1
0 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins