Application of amylomaltase for the synthesis of salicin-α-glucosides as efficient anticoagulant and anti-inflammatory agents

Article abstract of DOI:10.1016/j.carres.2016.06.011

The focus of this study was the synthesis of α-glucosyl derivatives of salicin by a transglucosylation reaction. The reaction was catalyzed by recombinant amylomaltase using tapioca starch as a glucosyl donor. Several reaction parameters, such as the enzyme-substrate concentrations, pH, temperature and incubation time, were optimized. Using the optimum conditions, at least three products with retention times (R_t) of 6.2, 9.2 and 14.1 were observed. The maximum yield of glucosylated salicin derivatives was 63% (w/w) of the total products. The structures of the glucosylated salicin derivatives were confirmed to be salicin-α-D-glucopyranoside, salicin-α-D-maltopyranoside and salicin-α-D-maltotriopyranoside through a combination of enzyme treatments, mass spectrometry and NMR analyses. The glycosidic bond between glucose units consisted of an α-1,4-configuration. The water solubility of salicin-α-D-glucopyranoside, salicin-α-D-maltopyranoside and salicin-α-D-maltotriopyranoside was 3-, 5- and 8-fold higher, respectively, than that of salicin, whereas their relative sweetness values were lower than that of sucrose. Interestingly, the long-chain salicin-α-D-glucosides showed greater anticoagulant and anti-inflammatory activities than salicin. In addition, the synthesized salicin-α-D-glucosides were able to tolerate acidic and high temperature conditions, but not α-glucosidase or human digestive enzymes. Therefore, these salicin-α-D-glucosides should be applied by the injection route to achieve greater bioavailability than is possible by the oral route.

Full text of DOI:10.1016/j.carres.2016.06.011

Carbohydrate Research 432 (2016) 55e61
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Application of amylomaltase for the synthesis of salicin-
a
-glucosides
as efficient anticoagulant and anti-inflammatory agents
,
*
Prakarn Rudeekulthamrong a, Jarunee Kaulpiboon b
a Department of Biochemistry, Phramongkutklao College of Medicine, Phramongkutklao Hospital, Bangkok 10400, Thailand
b Department of Pre-Clinical Science (Biochemistry), Faculty of Medicine, Thammasat University, Pathumthani 12120, Thailand
a r t i c l e i n f o
a b s t r a c t
Article history:
The focus of this study was the synthesis of
a
-glucosyl derivatives of salicin by a transglucosylation
Received 29 April 2016
Received in revised form
25 June 2016
Accepted 27 June 2016
Available online 29 June 2016
reaction. The reaction was catalyzed by recombinant amylomaltase using tapioca starch as a glucosyl
donor. Several reaction parameters, such as the enzyme-substrate concentrations, pH, temperature and
incubation time, were optimized. Using the optimum conditions, at least three products with retention
times (Rt) of 6.2, 9.2 and 14.1 were observed. The maximum yield of glucosylated salicin derivatives was
63% (w/w) of the total products. The structures of the glucosylated salicin derivatives were confirmed to
be salicin-
through a combination of enzyme treatments, mass spectrometry and NMR analyses. The glycosidic bond
between glucose units consisted of an -1,4-configuration. The water solubility of salicin- -D-gluco-
pyranoside, salicin- -D-maltopyranoside and salicin- -D-maltotriopyranoside was 3-, 5- and 8-fold
higher, respectively, than that of salicin, whereas their relative sweetness values were lower than that
of sucrose. Interestingly, the long-chain salicin- -D-glucosides showed greater anticoagulant and anti-
inflammatory activities than salicin. In addition, the synthesized salicin- -D-glucosides were able to
tolerate acidic and high temperature conditions, but not -glucosidase or human digestive enzymes.
Therefore, these salicin- -D-glucosides should be applied by the injection route to achieve greater
bioavailability than is possible by the oral route.
a-D-glucopyranoside, salicin-a-D-maltopyranoside and salicin-a-D-maltotriopyranoside
Keywords:
Amylomaltase
Salicin-a-D-glucosides
Anticoagulant agent
Anti-inflammatory agent
Tapioca starch
a
a
a
a
a
a
a
a
© 2016 Elsevier Ltd. All rights reserved.
1. Introduction
[7,8]. The first two activities are the main activities. Amylomaltase
has many potential applications. Firstly, it is used in the production
of LR-CDs via a cyclization reaction. An LR-CD can form a inclusion
complex with guest molecules by trapping them into a hydropho-
bic cavity, thus resulting in a change in the solubility, stability and
biological properties of the guest molecules [7]. Secondly, amylo-
maltase is used to modify starch and produce a thermo-reversible
starch gel with a gelatin-like property that can be used as a low
fat or vegetarian heavy cream substitute in dairy products [5,9,10].
Finally, the enzyme can be used in the production of linear oligo-
saccharides and glucoside products through an intermolecular
transglucosylation reaction. For example, short-chain iso-
maltooligosaccharides (IMOs), such as DP2-DP6, with prebiotic
activity were reported to be synthesized by a combination of
amylomaltase and transglucosidase [11].
Amylomaltase (EC 2.4.1.25), an intracellular 4-a-glucano-
transferase (4aGTase), was first identified in Escherichia coli as an
enzyme involved in maltose metabolism [1,2]. The enzyme was
subsequently reported to be expressed in most thermophilic and
archaea bacteria, such as Thermus aquaticus [3], Aquifex aeolicus [4]
and Pyrobaculum aerophilum IM2,5 and corresponded to the D-
enzyme in plants, such as potato [6]. Amylomaltase catalyzes an
intermolecular transglucosylation reaction, transferring glucosyl
residues from donor to acceptor as free OH groups of C4 to produce
longer linear oligosaccharides. In addition, this enzyme generates a
unique intramolecular transglucosylation reaction to yield cyclo-
amyloses or large-ring cyclodextrins (LR-CDs) with a degree of
polymerization (DP) of 16 or more. The enzyme has four different
activities: disproportionation, cyclization, coupling and hydrolysis
Salicin (C13H18O7: 2-hydroxymethyl-phenyl
b-D-glucopyrano-
side) is an alcoholic -glucoside extracted from several species of
b
Salix (willow) and Populus (poplar). In addition, salicin was also
found in Gaultheria procumbens (wintergreen) and in Betula lenta
(sweet birch), the volatile oils of which consist almost entirely of
* Corresponding author.
E-mail address: jkaulpiboon@yahoo.com (J. Kaulpiboon).
http://dx.doi.org/10.1016/j.carres.2016.06.011
0008-6215/© 2016 Elsevier Ltd. All rights reserved.

56
P. Rudeekulthamrong, J. Kaulpiboon / Carbohydrate Research 432 (2016) 55e61
methyl salicylate [12]. Salicin is used as an analgesic, anti-
inflammatory and antipyretic agent [13]. Glucosylation of salicin
has been the subject of increasing attention to improve its phar-
macokinetic parameters [14]. There are several reports on the
glucosylation of salicyl alcohol by cultured plant cells showing that
the benzyl hydroxyl group was glucosylated to yield isosalicin [15].
However, the reaction products were mostly synthesized by the
addition of a single glucose to the substrate. Until now, there have
been, to our knowledge, only two reports on the glucosylation of
salicyl alcohol or salicin using bacterial enzymes. In 2005, Seo et al.
[12] successfully modified salicyl alcohol, phenol and salicin using
Leuconostoc mesenteroides glucansucrase (EC 2.4.1.5) with sucrose
the intracellular crude amylomaltase fraction was obtained after
cell sonication. The enzyme was purified using a HisTrap FF™
column, as previously described [17]. The active fractions were
pooled and assayed for enzyme activity using a starch trans-
glucosylation assay [17], and the protein concentration was deter-
mined by the Bradford method [20] using BSA as the standard.
2.3. Donor specificity
The appropriate donor was selected by incubating 1% (w/v)
salicin and 40 U/ml amylomaltase in 0.2 M phosphate buffer at pH
6.0 with 1% (w/v) of the different glucosyl donors (starches and
maltoheptaose) in a 160 m
l reaction at 70 ꢀC for 24 h. The transfer
as a glucosyl donor. Salicin, phenyl glucose, isosalicin, salicyl
b-
isomaltooligosaccharides, and salicyl alcohol -iso-
b
products were analyzed using thin-layer chromatography (TLC)
with the mobile phase described in Section 2.5. The densities of the
salicin-a-glucoside spots on the TLC plate were measured using a
densitometer using the GS-800 program (Bio-Rad Laboratories, CA,
USA).
maltooligosaccharides were the major products. Later, Deinococcus
geothermalis (DGAS) and Neisseria polysaccharea (NPAS) amylosu-
crases (EC 2.4.1.4) were used to synthesize glucosyl and maltosyl
salicin from sucrose and salicin substrates [16]. Two salicin a-glu-
cosides from NPAS amylosucrase were detected and identified as
glucosyl and maltosyl salicin. In contrast, DGAS amylosucrase
synthesized only one glucosyl salicin product.
2.4. Optimization of the production of salicin-
Gn)
a-glucosides (b-Sa-a-
Hence, in this study, more attention has been paid to the syn-
thesis of a long carbohydrate salicin glucoside chain by amylo-
maltase using inexpensive, widely available substrates, such as
tapioca starch, which is a major economic crop in Thailand. In our
previous study, the amylomaltase gene was directly isolated from
soil DNA, cloned into a pET19b vector, and expressed in E. coli
BL21(DE3). The ORF of this gene consisted of 1572 bp, encoding an
enzyme of 523 amino acids [17]. Although it showed 99% sequence
identity to amylomaltase from Thermus thermophilus ATCC 33923,
this enzyme is unique regarding its alkaline optimum pH. The
enzyme produced cycloamyloses or LR-CD products in the range of
DP23- > DP50 through its intramolecular transglucosylation ac-
The optimum conditions for the production of salicin-
sides were considered and defined in terms of obtaining the highest
percentage yield of salicin- -glucoside products, as determined
a-gluco-
a
from the HPLC results. The effects of varying the salicin concen-
tration (0e2% (w/v)), tapioca starch concentration (0e2% (w/v)),
amylomaltase concentration (40e120 U/ml, starch trans-
glucosylation activity), pH (5.0e8.0), temperature (50e90 ꢀC), and
incubation time (0e29 h) were all investigated. After completion of
the incubation period, all tested reactions were inactivated by
boiling at 100 ꢀC for 10 min prior to HPLC analysis. The conversion
yield of glucosylated salicin was calculated from the ratio of the
tivity [18]. The prebiotic a-1,6-isomaltooligosaccharides were suc-
peak area of the salicin-
a-glucosides to the peak area of the total
cessfully synthesized from tapioca starch by the co-action of this
amylomaltase and transglucosidase through their intermolecular
transglucosylation reactions [19]. In this study, we focus on the
synthesis of new salicin derivatives using Thermus sp. amylo-
maltase with salicin and tapioca starch substrates in an effort to
increase the value of the agricultural product. In addition, a com-
parison of the inhibitory effects of the salicin analogs on blood
coagulation and inflammation is also performed.
products in the HPLC profile using equation (1):
peak area of
peak area of total product
b
ꢁ Sa ꢁ
a
ꢁ Gn
Conversion yield ð%Þ ¼
ꢂ 100
(1)
2.5. Product analysis
2. Materials and methods
The transglucosylation products were detected and identified
using HPLC and TLC analyses. For the HPLC analysis, the reaction
mixture of transglucosylation products was filtered using a 0.45-
2.1. Materials
mm filter and analyzed with a 5 mm SphereClone™ NH2 column
b-Salicin was purchased from Tokyo Chemical Industry Co., Ltd.
(4.6 ꢂ 250 mm) (Phenomenex, USA) connected to a Shimadzu LC-
9A system with a SPD-6AV UVeVIS detector at 265 nm [16]. The
column was eluted with 75% acetonitrile as the mobile phase using
a flow rate of 1.0 ml/min at room temperature. TLC was performed
with silica gel 60 F254 plates (Merck, Germany) after activation at
(Tokyo, Japan). D(þ)-Glucose, maltooligosaccharides (MOSs) with a
degree of polymerization (DP) of 2e7, potato starch, rat intestinal
acetone powder,
(BSA) were obtained from Sigma-Aldrich (USA). The Bio-Gel®P-2
Gel was obtained from Bio-Rad Laboratories (USA). A. niger
a-amylase, heparin and bovine serum albumin
a
-
95 ꢀC for 5 min. A 10-
onto a plate and developed with a solvent system of 5:5:3 (v/v/v) n-
butanol-ethanol-water in a TLC tank. Salicin- -glucosides were
ml aliquot of each reaction mixture was loaded
glucosidase (EC 3.2.1.20) was obtained from Amano Enzyme Inc.
(Nagoya, Japan). Tapioca starch was a gift from Siam Modified
Starch Co., Ltd. (Thailand). Rabbit blood without anticoagulant was
purchased from the National Laboratory Animal Center, Mahidol
University (Thailand). All other chemicals used were of analytical
grade.
a
detected by spraying the plates with a mixture of concentrated
sulfuric acid and methanol (1:2, v/v), and then heating them at
110 ꢀC for 20 min [21].
2.2. Amylomaltase preparation
2.6. Larger scale production of salicin-a-glucosides
Escherichia coli cells harboring the p19bAMY plasmid were
grown in LB medium containing 100
24 h. Enzyme expression was induced by adding 0.8 mM IPTG and
Larger-scale production was undertaken in a reaction volume of
48 ml using the optimized reaction conditions to obtain a greater
quantity of the salicin-a-glucoside products for the subsequent
m
g/ml ampicillin at 37 ꢀC for

P. Rudeekulthamrong, J. Kaulpiboon / Carbohydrate Research 432 (2016) 55e61
57
studies. Tapioca starch (2% (w/v)) was incubated with 120 U/ml
2.9.3. Acidic resistance
The acidic resistance of the salicin derivatives was investigated
in 100 mM acetate buffer, pH 3.0, containing 2% (w/v) salicin-
glucosides synthesized from amylomaltase or original salicin. The
reaction mixture was incubated at 37 ꢀC for 24 h. The remaining
salicin and salicin derivatives were determined by HPLC [23].
(starch transglucosylation unit) of amylomaltase in 0.2 M phos-
phate buffer at pH 6.0 and 70 ꢀC for 3 h. After completion, the re-
action mixture was stopped by boiling it for 10 min. The reaction
products were dried in a lyophilizer (Biotechnologies Inc., Ger-
many) before being applied to a Bio-Gel®P-2 Gel column.
a
-
2.9.4. Digestive enzyme treatment
The resistance of the products to digestive enzymes was tested
to evaluate the route of drug administration. One milligram of the
2.7. Purification of the salicin-
a
-glucosides
The mixture of dried salicin-
a
-glucoside products was dissolved
salicin-a-D-maltotriopyranoside product was mixed and dissolved
with distilled water and clarified by centrifugation before being
applied to a Bio-Gel®P-2 Gel column (1.2 ꢂ 97 cm). The column was
equilibrated and eluted with distilled water at a flow rate of 8 ml/h,
and 1-ml fractions were collected during this process. The carbo-
hydrates in the fractions were measured using the DNS assay [22].
The positive fractions were then analyzed by HPLC, and each pure
fraction was collected and lyophilized to analyze the products.
in 0.5 ml of 0.2 M acetate buffer, pH 5.5, followed by an incubation
with 0.5 ml of a-glucosidase (10 U/ml), a-amylase (10 U/ml) and rat
digestive enzymes (10 mg/ml, I1630-Sigma-Aldrich) at 37 ꢀC for 0,1
and 24 h. After incubation for each of these time intervals, the re-
action was halted by boiling and a 10-
then analyzed by TLC (System II).
ml aliquot of each reaction was
2.9.5. Anticoagulant activity
2.8. Structural analysis of salicin-a-glucosides
The whole blood clotting time was determined using the
method of Seo et al., [12] with slight modifications. This experiment
was performed in clean glass tubes immediately after blood was
drawn from a healthy rabbit of the National Laboratory Animal
Center, Mahidol University (Thailand). Then, 0.1 ml of a 10% (w/v)
solution of salicin and the salicin derivatives in phosphate buffer
(pH ¼ 7.4) was incubated with 1.4 ml of rabbit blood for 5,10,15 and
20 min. The time required for clotting was calculated immediately
after salicin and the salicin derivatives were added. If clotting was
not observed after 20 min, the sample was defined as not forming a
blood clot (NC). Heparin (0.5 and 50 U/ml) was used as a positive
control.
2.8.1. Mass spectrometry (MS) analysis
The transfer products of interest were dissolved in a 50% (v/v)
methanol solution, and their masses were determined on a
microTOF spectrometer (Bruker, Germany). The products were
introduced into a mass spectrometer for processing in the ESI-TOF-
MS system, which ionizes the sample by electrospray ionization
(ESI) in the sodium positive ion mode using a capillary voltage of
5000 V. A 4.0 L/min flow of nitrogen gas at a temperature of 150 ꢀC
was used to nebulize the analytic solution to droplets at a nebulizer
pressure of 1.0 bar. The ions were detected using linear time-of-
flight mass spectrometry (TOF-MS), and the mass to charge (m/z)
spectra of the products were detected as pseudo-molecular ion
peaks [MþNa]þ, from which the molecular weights were calculated
using a Bruker Daltonics Data Analysis 3.4 software program.
2.9.6.
b
-Glucuronidase inhibition assay
l of 2.5 mM p-nitrophenyl-b-D-glucopyr-
For this assay, 100
m
anosiduronic acid in 0.1 M acetate buffer (pH ¼ 7.4) was incubated
with 1 mg of salicin and the salicin derivatives for 5 min, followed
by the addition of 0.1 ml of b-Glucuronidase (Merck, France). The
2.8.2. Nuclear magnetic resonance (NMR) analysis
reaction mixture was then incubated for 30 min, followed by the
addition of 2 ml of 0.5 N NaOH to halt the reaction. The amount of
reaction product formed was measured at 410 nm and calculated
using equation (2):
The structures of the salicin-a-glucosides synthesized using
amylomaltase were identified with 1H- and 13C-NMR using a Bruker
AVANCE™ III HD 600 NMR Spectrometer and standard Bruker NMR
software. The spectrometer was operated at 600 MHz at ambient
temperature. Chemical shifts were measured with sodium-4,4-
dimethyl-4-sila-pentane sulfonite (DSS) as the internal standard.
The purified products (2 mg) were freeze-dried and dissolved in
1 ml of CD3OD containing 0.1% DSS prior to 1H- and 13C-NMR
analysis.
Anti ꢁ inflammatory activity ð%Þ
ð1 ꢁ A410 of test sampleÞ
¼
ꢂ 100
(2)
A410 of control sample
A 1 mM salicylic acid solution was used as a reference drug for
comparison [24].
2.9. Physical and biological properties of the salicin-
products
a-glucoside
2.10. Statistical analyses
2.9.1. Water solubility
All data are expressed as the means SD from at least three
separate experiments, and the differences were calculated using
Student's t-test (GraphPad Prism software version 4). P-values <
0.05 were considered statistically significant.
Excess salicin and salicin-a-glucosides were mixed with 200 ml
of water in micro-centrifuge tubes at room temperature. After
20 min of mixing with a vortex mixer at 25 ꢀC, each of the samples
was diluted and filtered through a 0.45-mm membrane for HPLC
analysis to determine the concentrations.
3. Results and discussion
2.9.2. Sweetness
3.1. Donor specificity
The relative sweetness of 1% (w/v) of the tested salicin-a-glu-
cosides was determined by measuring the Brix values using a
refractometer (RX-1000, Atago Co., Ltd., Tokyo, Japan) [23]. Sucrose
was used as the reference compound to plot the standard curve of
the concentrations (%, w/v) versus the degrees Brix (ꢀBx). One de-
gree Brix is defined as 1 g of sucrose in 100 ml of solution.
A 1% (w/v) salicin solution was used as the acceptor and 1% (w/v)
potato starch, tapioca starch and maltoheptaose (G7) were used as
glucosyl donors in a transglucosylation reaction with recombinant
amylomaltase (40 U/ml) in 0.2 M phosphate buffer at pH 6.0 at
70 ꢀC for 24 h to determine the potency of different glucosyl donors

58
P. Rudeekulthamrong, J. Kaulpiboon / Carbohydrate Research 432 (2016) 55e61
for salicin. The products were analyzed by TLC. Upon analysis of the
product by TLC, at least three products with higher Rf values than
the sugar standard were observed (Fig. S1). These trans-
(Fig. S2). Under these conditions, the yield of the salicin-a-gluco-
side products was up to 63% (w/w) (18.9 mg/ml) of the total
products, which were increased more than 1.8-fold compared to
the yield before optimization. This increase in product yield rep-
glucosylation products were believed to be the salicin-a-glucosides,
which were synthesized by an amylomaltase-catalyzed trans-
glucosylation reaction from tapioca starch to the salicin acceptor. In
addition, the transglucosylation yields were also determined from
resented an increase in all salicin-
ucts. The concentrations of -Sa-
were 9.9, 5.4 and 3.6 mg/ml, respectively, compared with the
standard curve of each salicin- -glucoside. In general, the synthesis
of salicin glucosides by the action of other bacterial enzymes has
been reported. For example, the incubation of Leuconostoc mesen-
teroides 1299 CB-BF563 glucansucrases with sucrose and salicin
a
-glucoside,
b
-Sa-
a
-G1-3, prod-
b
a
-G1, -Sa-
b
a
-G2 and
b
-Sa- -G3
a
the relative ratio of the salicin-
product-spot density. The salicin-
a
-glucoside-spot density to total
-glucoside yields of potato
a
a
starch, tapioca starch and G7 were 36, 36 and 20% (w/w) of the total
products, respectively. Based on this experiment, tapioca starch, an
inexpensive and widely available substrate, was chosen as a suit-
able glucosyl donor for transglucosylation to a salicin acceptor.
could catalyze the synthesis of salicyl alcohol
b-iso-
maltooligosaccharides, with yields of 47% (w/w) [12]. The optimum
conditions for generation of the salicin transfer product by the
Neisseria polysaccharea (rNPAS) and Deinococcus geothermalis
(rDGAS) amylosucrases with salicin and sucrose substrates were
35 ꢀC (pH ¼ 6) and 45 ꢀC (pH ¼ 8), respectively [16]. Their HPLC
analyses showed that two glucosylated salicin products (designated
as glucosyl and maltosyl salicin) were generated in the reaction
with the rNPAS enzyme, but only one product (glucosyl salicin) was
generated in the reaction with the rDGAS enzyme. The yield of the
salicin transfer products was over 80% in the rNPAS reaction, but
the yield of the rDGAS reaction (70%) was lower than the rNPAS
3.2. Optimization of salicin-a-glucoside synthesis
The parameters of the reaction were sequentially and inde-
pendently optimized by changing some of the conditions of the
transglucosylation reaction to produce the highest yield of the
salicin-a-glucosides. These parameters included the substrate
concentrations, enzyme concentrations, pH, temperature and in-
cubation time. The products obtained from each reaction were
analyzed by HPLC and TLC, compared to the salicin standard (Fig. 1),
and quantified using equation (1). As a result, the optimum con-
reaction. Accordingly, the salicin-a-glucoside yield obtained in this
study was quite sizeable as compared to the group involved in the
bacterial enzyme-catalyzed synthesis. Specifically, we learned that
our amylomaltase is a unique enzyme that acts differently than
dition for the synthesis of salicin-a-glucosides was an incubation of
1% (w/v) salicin and 2% (w/v) tapioca starch with 120 U/ml amy-
lomaltase in 0.2 M phosphate buffer at pH 6.0 and 70 ꢀC for 3 h
Fig. 1. HPLC and TLC analyses of the products from the reaction of recombinant amylomaltase with the salicin acceptor and tapioca starch donor. (A.) HPLC analysis: (a.) reaction at
0 h and (b.) reaction at 3 h under the optimal conditions. (B.) TLC analysis: Lane M - G1-G7 standards, Lane 1e2% (w/v) tapioca starch, Lane 2 - 1% (w/v) salicin, Lanes 3 and 4 -
products from the enzymatic reaction at 0 and 24 h, Lane 5 - salicin (HPLC peak I), Lane 6 - purified product 1 (HPLC peak II), Lane 7 - purified product 2 (HPLC peak III), and Lane 8 -
purified product 3 (HPLC peak IV).

P. Rudeekulthamrong, J. Kaulpiboon / Carbohydrate Research 432 (2016) 55e61
59
Table 2
other enzymes in the transglycosylation of oligosaccharides to
salicin for the synthesis of novel salicin- -D-maltotriopyranosides.
1H-NMR chemical shift values and coupling constants for the glucosidic protons of
salicin and the salicin transglucosylation products.
a
Compound
-Salicin (
Product 1 (
Product 2 (
Product 3 (
Glucosidic protons
3.3. Large scale preparation, purification and structural analysis of
the synthesized salicin- -glucosides
b
b
-Sa)a
-Sa-
-Sa-
-Sa-
4.86b (d,8.0)c
-G1) 4.88 (dd,8.0) 5.23 (d,4.0)
a
b
b
b
a
a
a
-G2) 4.87 (d,8.0)
-G3) 4.87 (d,8.0)
5.18 (d,4.0) 5.24 (d,4.0)
5.18 (d,4.0) 5.25 (d,4.0) 5.28 (d,4.0)
The reaction mixture was incubated under the optimal condi-
tions and scaled up to 48 ml to prepare a larger amount of the
synthesized salicin-a-glucosides for the structural analysis. The
reaction mixture was then concentrated with a lyophilizer. The
concentrated reaction products were loaded onto a Bio-Gel-P2
column and separately analyzed by HPLC before mass analysis.
The molecular weights of the purified products were confirmed
with ESI-TOF MS in the positive mode. There were intense signals at
m/z 471.28 (Product 1), 633.25 (Product 2) and 795.38 (Product 3),
respectively (Fig. S3). These three reaction products corresponded
a
Jung et al. [16].
Chemical shifts (d) were given in ppm.
The coupling constants (Hz) are presented in parentheses.
Doublet.
b
c
d
between glucose I and II. The type of glucosidic linkage was also
investigated using 1H-NMR and was based on the coupling con-
stants of the anomeric protons, as shown in Table 2 b-Sa-a-G1 gave
a double signal at 4.88 and 5.23 ppm, with coupling constants (J) of
8.0 and 4 Hz, respectively, which suggested that the type of
glucosidic linkage was one
tion. When -Sa- -G1 was analyzed in terms of its structure and in
combination with its molecular weight, we concluded that the
to the molecular weight of salicin-
[16], salicin- -D-maltopyranoside (
maltotriopyranoside ( -Sa- -G3), respectively.
The structures of the salicin- -glucosides were also analyzed by
13C- and 1H-NMR to identify the structure of salicin glucoside
products, as shown in Tables 1 and 2 (Figs. S4 and S5). For -Sa- -G1
a-D-glucopyranoside (
b-Sa-
a-G1)
a
b
-Sa- -G2) [16] and salicin-
a
a-D-
b-configuration and one a-configura-
b
a
b
a
a
b
-
Sa-
a
-G1 product was salicin-
a
-D-glucopyranoside (Fig. 2A).
-Sa- -G2 (Product 2) and
-G3 (Product 3) and the analysis of the 13C-NMR spectra of
-G2 and -Sa-
b
a
Likewise, the structural analysis of
a
a
b
a
b
b
-
-
(Product 1; Rt ¼ 6.2 min), the 13C-NMR analysis displayed nineteen
carbon signals. Seven were assigned as salicyl alcohol. Another
twelve signals were assigned to the glucosyl group. The C40 signal of
the glucose I unit (Product 1) changed greatly from 71.34 ppm
(Salicin) to 80.90 ppm (þ9.56 ppm) because of the glucosylation
Sa-
Sa-
b
a
-G3 were also performed and revealed twenty-
five and thirty-one carbon signals, respectively (Table 1). Thirteen
of the signals were assigned to the salicin group, and the remaining
twelve and eighteen signals were construed as two and three units
of glucose. Compared to the C400-glucose II signals of Product 1, the
C400-glucose II signals of Product 2, and C400-glucose II and C4000-
glucose III signals of Product 3 showed downfield chemical shifts of
9.82, 9.79 and 9.89 ppm, respectively. These downfield shifts
Table 1
13C-NMR analysis of salicin and the salicin transglucosylation products.
Carbon atom
13C-NMR (
Salicina
d
, ppm)
indicated that b-Sa-a-G2 (Product 2) and b-Sa-a-G3 (Product 3) had
Product 1
Product 2
Product 3
two and three glucosidic linkages with salicin, respectively. One of
the glucosidic linkage was between the salicin group and glucose,
and the other was between glucose and glucose. In addition, the
C-salicyl alcohol
1
2
3
4
5
6
7
156.99
132.03
129.75
123.62
129.85
116.96
60.98
157.06
132.21
129.92
123.80
129.97
117.04
60.98
157.06
132.22
129.91
123.81
129.97
117.06
60.98
157.09
132.29
129.93
123.83
129.95
117.14
60.99
1H-NMR spectrum of
configuration and two
signals at 4.87, 5.18 and 5.24 ppm (J ¼ 8.0, 4.0 and 4.0 Hz). Similarly,
the structure of -Sa- -G3 was one -configuration and three a-
b
-Sa-a-G2 revealed that b-Sa-a-G2 had one b-
a
-anomeric configurations, based on the
b
a
b
configurations that were confirmed by the signals of the anomeric
proton at 4.87, 5.18, 5.25 and 5.28 ppm (J ¼ 8.0, 4.0, 4.0 and 4.0 Hz,
respectively) by 1H-NMR analysis (Table 2). During the course of the
integrated NMR and MS analysis, Products 2 and 3 were found to be
C-glucose I
10
103.29
75.05
77.96
71.34
78.21
62.53
103.19
74.70
77.75
80.90
76.86
62.04
103.13
74.70
77.73
80.92
76.85
62.08
103.27
74.73
77.77
80.86
76.89
62.13
20
30
40
salicin-a-D-maltopyranoside and salicin-a-D-maltotriopyranoside
50
(Fig. 2B and C).
60
C-glucose II
100
102.94
74.19
75.11
71.56
74.84
62.73
102.94
73.81
74.97
81.38
73.41
62.19
102.89
73.81
74.96
81.35
73.41
62.20
3.4. Physical and biological properties of the salicin-
products
a-glucoside
200
300
400
500
3.4.1. Water solubility
600
The solubility of each of salicin-a-glucoside was compared to
C-glucose III
that of the original salicin. Excess salicin and salicin glucosides were
mixed with water and incubated at 25 ꢀC for 20 min. The soluble
part of the sample was analyzed by HPLC to determine the con-
centrations. Based on the result, the solubility of salicin was
1000
102.71
74.28
75.13
71.53
74.81
62.77
102.74
74.30
75.15
81.45
74.80
62.80
2000
3000
4000
5000
25.42 mg/ml, whereas the solubilities of
b-Sa-a-G1, b-Sa-a-G2 and
6000
b
-Sa- -G3 were 74.32, 132.36 and 209.57 mg/ml, respectively, cor-
a
C-glucose IV
10000
20000
30000
40000
50000
60000
102.65
73.43
75.00
71.59
73.92
62.25
responding to levels that were 3, 5 and 8 times higher than that of
salicin. This result implied that the glucosylation process could be
used to increase the solubility of salicin.
3.4.2. Sweetness
a
Jung et al. [16].
The sweetness values of salicin and the synthesized b-Sa-a-G1,

60
P. Rudeekulthamrong, J. Kaulpiboon / Carbohydrate Research 432 (2016) 55e61
Fig. 2. Structures of salicin and the salicin-a-D-glucosides. A. Salicin-a-D-glucopyranoside. B. Salicin-a-D-maltopyranoside. C. Salicin-a-D-maltotriopyranoside.
b
-Sa-
a
-G2 and
b
-Sa-
a
-G3 products were 0.25, 0.20, 0.1 and 0.05,
between salicyl alcohol and glucose, could not be digested with
either enzyme. This result is related to the fact that the amylo-
respectively, compared to the sweetness value of sucrose, which is
set to 1.0. Based on these results, the sweetness decreased as the
saccharide chain lengths of the salicin glucosides increased. Prac-
tically, the sweetness depends on the chemical structure, the de-
gree of polymerization of the oligosaccharides present, and the
levels of mono- and di-saccharides in the mixture [25]. According
to Roberfroid and Slavin [26], the sweetness of an oligosaccharide
maltase is specific for the
a-1,4-glucosidic bond [28].
3.4.5. Anticoagulant activity
The time required for clotting was measured to determine the
anticoagulant effects of salicin, the salicin derivatives, and aspirin
(Table S1). Aspirin and salicin produced blood clots within 5 min,
decreases with longer oligosaccharide chain lengths.
A low
whereas
15 min
b
-Sa-
a
-G1 and
b-Sa-a-G2 caused blood clots to form within
sweetness level is quite useful when the glucosides are used as a
sucrose substitute in a functional material where the use of sucrose
is restricted by its high sweetness property [14,27].
b-Sa-
a
-G3 formed blood clots within 20 min under the
described conditions. In addition, we also checked the anticoagu-
lant effect of heparin as a positive control. Heparin (0.5 U/ml)
produced slow blood clotting within 5 min and complete blood
clotting within 10 min, whereas heparin (50 U/ml) did not form
blood clots, even after 20 min had passed. Typically, the typical
dose of heparin that is used to treat and prevent blood clots in the
veins and arteries in the lungs is a final concentration of ꢃ0.5 U/ml
blood [29,30]. This result suggests that a dose of 0.67% (w/v) of the
salicin derivatives has a greater potential to inhibit the formation of
blood clots than 0.5 U/ml of heparin. The anticoagulant action of
3.4.3. Acidic resistance
Salicin and all salicin derivatives showed low degradability in an
acidic buffer at pH 3.0 after an incubation at 37 ꢀC for 24 h (Fig. S6).
Their high stabilities under acidic conditions suggests that the
synthesized salicin-a-glucosides could resist the gastric acidity of
humans and could be used in acidic materials, such as fruit juices,
alcoholic beverages, soft drinks, sparkling water and acidic drugs.
the salicin-a-glucosides in this study might be the result of the
inhibition of cyclooxygenase (COX) enzymes [31], which are
involved in thromboxane synthesis in the blood coagulation
system.
3.4.4. Digestive enzyme treatment
The route of drug administration was preliminarily investigated
by a digestive enzyme treatment, as described in Section 2.9.3. After
a 1-h treatment,
digestive enzymes, such as
amylase, sucrase, among others, yielding
After a longer treatment period of 24 h, glucose and salicin products
were observed (Fig. S7). This result suggested that salicin de-
rivatives should be introduced into the body by intravenous or
subcutaneous routes rather than by an oral route to avoid the
digestive tract. In addition, this result also confirmed that each
b
-Sa-
a
-G3 was degraded by
-glucosidase, maltase, isomaltase,
-Sa- -G1 and -Sa- -G2.
a-glucosidase and rat
a
a
-
3.4.6.
Anti-inflammatory activity was determined using the
ronidase inhibition assay. Salicin, -Sa- -G1, -Sa- -G2, and
G3 showed moderate anti-inflammatory activity measuring 15, 28,
32 and 40% of salicylic acid (control ¼ 100%), respectively. The re-
sults also showed that the long carbohydrate chains of salicin were
b-Glucuronidase inhibition assay
b
a
b
a
b
b
-glucu-
-Sa-a-
b
a
b
a
more effective at inhibiting
b-glucuronidase than shorter carbo-
glucose was linked with salicin through an
it easier to digest with -glucosidase and rat digestive enzymes. In
contrast, original salicin, which is linked by a -configuration
a
-configuration, making
hydrate chains. It is known that b
in lysosomes of neutrophils and plays an important role as medi-
ator of the initiation and progression of inflammation [32]. This
-glucuronidase is mainly located
a
b

P. Rudeekulthamrong, J. Kaulpiboon / Carbohydrate Research 432 (2016) 55e61
910e915.
61
result might explain how the modified salicin inhibited b-glucu-
[4] S.H. Bhuiyan, M. Kitaoka, K. Hayashi, J. Mol. Catal. B Enzym 22 (2003) 45e53.
[5] T. Kaper, B. Talik, T.J. Ettema, B. Herman, M.J.E.C. van der Maarel, L. Dijkhuizen,
Appl. Environ. Microbiol. 71 (2005) 5098e5106.
ronidase and how it is related to the inhibition of other inflam-
matory mediators, such as prostaglandins, lipoxins and
leukotrienes [32].
[6] T. Takaha, M. Yanase, H. Takata, S. Okada, S.M. Smith, J. Biol. Chem. 271 (1996)
2902e2908.
[7] T. Takaha, S.M. Smith, Biotechnol. Genet. Eng. Rev. 16 (1999) 257e280.
[8] T. Endo, M. Zheng, W. Zimmermann, Aust. J. Chem. 55 (2002) 39e48.
[9] M.J.E.C. van der Maarel, I. Caprona, G.J.W. Euverinka, H.T. Bosa, T. Kaper,
D.J. Binnema, et al., Starch 57 (2005) 465e472.
[10] A.C. Alting, F. van de Veld, M.W. Kanning, M. Burgering, L. Mulleners, A. Sein,
et al., Food Hydrocoll. 23 (2009) 980e987.
4. Conclusions
The enzyme from the p19bAMY recombinant cells containing
the amylomaltase gene from Thermus sp. could be used to syn-
thesize salicin-a-D-glucosides by transferring one or more glucose
[11] P. Rudeekulthamrong, K. Sawasdee, J. Kaulpiboon, Biotechnol. Bioprocess Eng.
18 (2013) 778e786.
[12] E.S. Seo, J.H. Lee, J.Y. Park, D. Kim, H.J. Han, J.F. Robyt, J. Biotechnol. 117 (2005)
31e38.
[13] T. Akao, T. Yoshino, K. Kobashi, M. Hattori, Planta Med. 68 (2002) 714e718.
[14] M. Schottelius, H.J. Wester, J.C. Reubi, R. Senekowitsch-Schmidtke,
M. Schwaiger, Bioconjug. Chem. 13 (2002) 1021e1030.
[15] A. Syahrani, I. Widjaja, G. Indrayanto, A.L. Wilkins, J. Asian Nat. Prod. Res. 1
(1998) 111e117.
residues from the glucosyl donor, tapioca starch, to the salicin
acceptor through the transglucosylation reaction. Under the
optimal conditions, the yield of the obtained salicin glucosides was
63% (w/w) of the total products. The acceptor products of salicin
were confirmed to be salicin-a-D-glucopyranoside, salicin-a-D-
maltopyranoside and salicin- -D-maltotriopyranoside by ESI-TOF
a
[16] J.H. Jung, D.H. Seo, S.J. Ha, M.C. Song, J. Cha, S.H. Yoo, et al., Carbohydr. Res. 344
(2009) 1612e1619.
[17] S. Watanasatitarpa, P. Rudeekulthamrong, K. Krusong, W. Srisimarat,
W. Zimmermann, P. Pongsawasdi, J. Kaulpiboon, Appl. Biochem. Microbiol. 50
(2014) 243e252.
MS and NMR analyses. Moreover, the products obtained from this
study also exhibited properties of a potential anticoagulant and
anti-inflammatory agent.
[18] K. Sawasdee, P. Rudeekulthamrong, W. Zimmermann, S. Murakami,
P. Pongsawasdi, J. Kaulpiboon, Appl. Biochem. Microbiol. 50 (2014) 17e24.
[19] J. Kaulpiboon, P. Rudeekulthamrong, S. Watanasatitarpa, K. Ito,
P. Pongsawasdi, J. Mol. Catal. B Enzym 120 (2015) 127e135.
[20] M.M. Bradford, Anal. Biochem. 72 (1976) 248e254.
[21] M. Funamaya, T. Nishimo, A. Hirota, S. Murao, S. Takenishi, H. Nakano, Biosci.
Biotech. Biochem. 57 (1993) 1666e1669.
Acknowledgements
The authors gratefully acknowledge the financial support pro-
vided by the Faculty of Medicine, Thammasat University under the
research fund Contract No. TP-2-02/2015. In addition, partial sup-
port from the Phramongkutklao College of Medicine Research Fund
is also acknowledged.
[22] P. Bernfeld, Amylase ∞ and b, in: S.P. Colowick, N.O. Kaplan (Eds.), Methods in
Enzymology, Vol. 1, Academic Press, New York, USA, 1955, pp. 149e158.
[23] H. Kikuchi, T. Nagura, M. Inoue, T. Kishida, H. Sakurai, A. Yokota, et al., J. Appl.
Glycosci. 51 (2004) 291e296.
[24] R. Nia, D.H. Paper, E.E. Essien, O.H. Oladimeji, K.C. Iyadi, G. Franz, Niger. J.
Physiol. Sci. 18 (2003) 39e43.
Appendix A. Supplementary data
[25] R.G. Crittenden, M.J. Playne, Trends Food Sci. Technol. 7 (1996) 353e361.
[26] M. Roberfroid, J. Slavin, Crit. Rev. Food Sci. Nutr. 40 (2000) 461e480.
[27] F. Schweizer, O. Hindsgaul, Curr. Opin. Chem. Biol. 3 (1999) 291e298.
[28] T. Takaha, S.M. Smith, Biotechnol. Genet. Eng. Rev. 16 (1999) 257e280.
[29] C. Kearon, J.S. Ginsberg, J.A. Julian, J. Douketis, S. Solymoss, P. Ockelford, et al.,
JAMA 296 (2006) 935e942.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.carres.2016.06.011.
References
[30] M. Colucci, A. Pentimone, B.M. Binetti, M. Cramarossa, D. Piro, N. Semeraro,
J. Thromb. Haemost. 88 (2002) 282e287.
[31] K. Schrӧr, Semin. Thromb. Hemost. 23 (1997) 349e356.
[32] S.H. Nile, C.N. Khobragade, Nat. Prod. Comm. 6 (2011) 1437e1440.
[1] J. Monod, A.M. Torriani, Ann. Inst. Pasteur (Paris) 76 (1950) 65e77.
[2] M. Doudoroff, W.Z. Hassid, E.W. Putman, A.L. Potter, J. Biol. Chem. 179 (1949)
921e934.
[3] Y. Terada, K. Fujii, T. Takaha, S. Okada, Appl. Environ. Microbiol. 65 (1999)