Transglycosylation Forms Novel Glycoside Ethyl α-Maltoside and Ethyl α-Isomaltoside in Sake during the Brewing Process by α-Glucosidase A of Aspergillus oryzae

Article abstract of DOI:10.1021/acs.jafc.9b06936

Sake, the Japanese rice wine, contains a variety of oligosaccharides and glucosides produced by fungal enzymes during the brewing process. This study investigates the effect of knocking out the Aspergillus oryzae α-glucosidase (agdA) gene on the transglycosylation products in brewed sake. In addition to α-ethyl glucoside and α-glyceryl glucoside, the amount of two compounds that have molecular mass values similar to that of ethyl maltose decreased by agdA gene knockout. Both compounds were synthesized, in vitro, from maltose and ethanol with purified agdA. Nuclear magnetic resonance analysis identified the two compounds as ethyl α-maltoside and ethyl α-isomaltoside, respectively, which are novel compounds in sake as well as in the natural environment. Quantitative analysis of 111 commercially available types of sake showed that these novel compounds were widely present at concentrations of several hundred mg/L, suggesting that both of them are ones of the common glycosides in sake.

Full text of DOI:10.1021/acs.jafc.9b06936

Subscriber access provided by BIU Pharmacie | Faculté de Pharmacie, Université Paris V
Food and Beverage Chemistry/Biochemistry
Transglycosylation forms novel glycosides ethyl #-
maltoside and ethyl #-isomaltoside in sake during the
brewing process by #-glucosidase A of Aspergillus oryzae
Yusuke Kojima, Chihiro Honda, Izumi Kobayashi, Ryo Katsuta, Satomi Matsumura, Izumi Wagatsuma,
Maya Takehisa, Hitoshi Shindo, Masaru Hosaka, Tomoo Nukada, and Masafumi Tokuoka
J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.9b06936 • Publication Date (Web): 30 Dec 2019
Downloaded from pubs.acs.org on January 4, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 25
Journal of Agricultural and Food Chemistry
Transglycosylation forms novel glycosides ethyl a-maltoside and ethyl a-isomaltoside in sake during the
brewing process by a-glucosidase A of Aspergillus oryzae
Yusuke Kojima1, Chihiro Honda1, Izumi Kobayashi1, Ryo Katsuta1, 3, Satomi Matsumura2, Izumi Wagatsuma1, Maya
Takehisa2, Hitoshi Shindo1, Masaru Hosaka1, Tomoo Nukada3, Masafumi Tokuoka1*
1Tokyo University of Agriculture, Graduate School of Agriculture, Department of Fermentation Science and
Technology, 1-1-1 Sakuragaoka, Setagaya-ku, Tokyo 156-8502, Japan
2Tokyo University of Agriculture, Department of Fermentation Science, 1-1-1 Sakuragaoka, Setagaya-ku, Tokyo
156-8502, Japan
3Tokyo University of Agriculture, Department of Chemistry for Life Sciences and Agriculture, 1-1-1 Sakuragaoka,
Setagaya-ku, Tokyo 156-8502, Japan
Corresponding author’s E-mail: m3tokuok@nodai.ac.jp
1
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 2 of 25
1
2
Abstract
Sake, the Japanese rice wine, contains a variety of oligosaccharides and glucosides produced by fungal enzymes
during the brewing process. This study investigates the effect of knocking out the Aspergillus oryzae a-glucosidase
(agdA) gene on the transglycosylation products in a brewed sake. In addition to a-ethyl glucoside and a-glyceryl
glucoside, the amount of two compounds that have molecular masses similar to ethyl maltose decreased by agdA
gene knockout. Both compounds were synthesized, in vitro, from maltose and ethanol with purified agdA. Nuclear
magnetic resonance analysis identified the two compounds as ethyl a-maltoside and ethyl a-isomaltoside,
respectively, which are novel compounds in sake as well as in the natural environment. Quantitative analysis of 111
commercially available types of sake showed that these novel compounds were widely present at concentrations of
several hundred mg/L, suggesting that both of them are ones of the common glycosides in sake.
3
4
5
6
7
8
9
10
11
12
13
Keywords: a-glucosidase, transglycosylation, glucoside, sake, brewing, Aspergillus oryzae
2
ACS Paragon Plus Environment

Page 3 of 25
Journal of Agricultural and Food Chemistry
14
INTRODUCTION
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Sake, a Japanese rice wine, is brewed with the “multiple parallel fermentation” process. Enzymes produced by the
fungus Aspergillus oryzae carry out the saccharification process while the yeast Saccharomyces cerevisiae
ferments alcohol simultaneously. During the sake-brewing process, fungal hydrolytic enzymes degrade the starch
of rice into glucose, which the yeast subsequently uses as the substrate for alcoholic fermentation. Glucose is also
an important component of sake that provides sweetness and harmony with other components, which form sake’s
taste. In general, glucose is the most abundant non-volatile component in sake, ranging in concentration from 1%
to 4%, followed by glycerol and isomaltose (6-O-a-D-glucopyranosyl-a,b-D-glucopyranose) 1-3. Yeast produces
glycerol, which is also present in wine and beer, as a by-product of alcoholic fermentation 4. Isomaltose is a product
of transglycosylation by fungal enzymes, but yeast does not preferentially assimilate it when glucose is present in
the culture medium 5. Other than isomaltose, kojibiose (2-O-a-D-glucopyranosyl-a,b-D-glucopyranose) and
sakebiose (3-O-a-D-glucopyranosyl-a,b-D-glucopyranose) 5 are generated by transglycosylation by fungal
enzymes, suggesting active transglycosylation during the sake-brewing process 5. Yeast does not use the
transglycosylation products in sake during alcoholic fermentation. Besides oligosaccharides, ethyl a-glucoside (a-
EG) and glyceryl a-glucoside (a-GG), which transglycosylation generates during the sake-brewing process, are
29
30
31
32
33
34
35
36
37
present in sake at concentrations of 0.24%–0.71% and 0.45%, respectively 6, 7. a-EG adds both sweetness and
bitterness to the unique taste of sake 6. Both a-EG and a-GG have skin moisturizing effects and, thus, are used in
the production of cosmetics 8-10. These glycosides are thought to be synthesized by transglycosylation of a
glycosidase only when an acceptor compound (e.g., ethanol or glycerol) and donor maltose are both present,
which is a unique situation in the multiple parallel fermentation process 11, 12
.
Transglycosylation may be catalyzed by glycosidase derived from A. oryzae 11, 13. However, the
glycosidase(s) responsible for generating such transglycosylation products in the sake-brewing process remain
unclear. The A. oryzae genome codes for several glucosidases, including a-glucosidase A (AgdA), which is
encoded by agdA 14. AgdA seems to be a major a-glucosidase because it is highly induced by starch (or maltose),
38
39
40
41
42
43
44
which is also observed in other amylolytic enzymes, such as a-amylase, encoded by amyA, B, and C, and
glucoamylase, encoded by glaA 15. Nevertheless, AgdA’s involvement in transglycosylation during the sake-
brewing process is unknown.
In this study, the agdA gene of A. oryzae was knocked out to identify which components of sake are
produced by AgdA’s enzymatic activity. The results identified two novel transglycosylation products (ethyl a-
maltoside and isomaltoside) that are generally present in sake.
3
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 4 of 25
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
MATERIALS AND METHODS
Chemicals
All reagents were analytical grade, unless otherwise stated, and purchased from FUJIFILM Wako Pure Chemical
Corporation (Osaka, Japan).
Construction of an agdA gene disruption (DagdA) DNA cassette
A DNA fragment containing the genomic region of the agdA open reading frame (ORF) and 500-bp flanking regions
were amplified by polymerase chain reaction (PCR) using Platinum™ Taq DNA Polymerase (Thermo Fisher
Scientific K.K., Tokyo, Japan), and cloned using TOPO TA Cloning® Kits (Thermo Fisher Scientific K.K.). The
resulting plasmid, pTAagdA, was linearized by PCR using KOD-Plus-Neo DNA polymerase (Toyobo Co., Ltd.,
Osaka, Japan). The selective marker gene sC, encoding sulfate adenylyltransferase, was amplified from the
plasmid pUSC 16 by PCR using KOD-Plus-Neo, and introduced into the linearized plasmid pTAagdA using the
InFusion cloning kit (Takara Bio, Inc., Shiga, Japan), generating the plasmid pKSagdA. The DNA cassette for
DagdA was amplified from pKSagdA. Table S1 lists all the primers used for PCR. Plasmid propagation was
performed with ECOS™ Competent E. coli DH5a (Toyobo Co., Ltd.).
Construction of an AgdA overexpression DNA cassette
A fragment of the agdA ORF, amplified by PCR from the genomic DNA of A. oryzae strain RIB40, was inserted
downstream of the modified enolase A promoter sequence of the PCR-linearized pNENU2512 vector 17 by InFusion
cloning. The resulting AgdA overexpression vector pEAGN was used for fungal transformation.
Fungal transformation
In order to generate the DagdA strain, A. oryzae strain NS4DligD 18 was transformed with a PCR fragment amplified
from the plasmid pKSagdA. Transformation was performed using a standard protoplast method 19. Czapek–Dox
agar (3% glucose, 0.1% K2HPO4, 0.05% KCl, 0.05% MgSO4・7H2O, 0.001% FeSO4, and 1.5% agar), supplemented
70
71
72
73
74
75
with 0.9% (NH4)2SO4 as a nitrogen source, was used as a selective medium. A. oryzae strain NS4DligD was
transformed by PCR-amplification of a DNA fragment of sC to generate the control strain NSCS. A. oryzae strain
niaD300 16 was transformed with the plasmid pEAGN digested with BamHI to generate the AgdA overexpression
strain EAGN. Except for addition of 20 U of BamHI into the protoplast and DNA mixture, transformation was
performed using a standard protoplast isolation method 19 with Czapek–Dox agar as the selective medium.
4
ACS Paragon Plus Environment

Page 5 of 25
Journal of Agricultural and Food Chemistry
76
Southern blot analysis
77
78
79
80
81
82
83
84
Genomic DNA extraction, hybridization, and detection were performed as described previously 20. Genomic DNA of
the candidate DagdA strain and the host strain were digested with SalI, while that of the EAGN and host strains
were digested with HindIII. An 877-bp region containing the promoter and ORF of agdA and a 2.0-kb region
containing the promoter and ORF of pEAGN were used to generate a digoxigenin-labeled probe.
Small-scale sake brewing and evaluation
Koji was prepared in accordance with the method described by Tokuoka et al. 21 using 70% polished ordinary rice
and the conidia of strains DagdA and NSCS. Small-scale brewing was performed with ordinary rice (70% polished),
85
86
87
88
89
90
91
92
93
rice-koji produced by strains DagdA and NSCS, and yeast (S. cerevisiae K-901) using a three-step mashing
processes. As the first step (Soe), 24 g of rice and 12 g of rice-koji were added to 60 mL of water containing yeast
at 106 cells/mL. The next day, as the second step (Naka), 52 g of rice, 12 g of rice-koji, and 92 mL of water were
added to the fermentation mash. On day 3, as the third step (Tome), 84 g of rice, 16 g of rice-koji, and 128 mL of
water were added. The mash was fermented at 15°C for 20 days and then centrifuged at 2,000 x g for 10 min to
separate the sake cake from the sake.
The sake meter value (a measure of the density of sake relative to water), alcohol content, and total acid
content were measured with the use of conventional methods 22. The amount of fungal cells was estimated based
on the amount of N-acetyl glucosamine in accordance with the methods described by Reissig et al. 23 and Fujii et
94
95
96
97
98
99
al. 24
.
Analysis of a -amylase and a -glucosidase
Rice-koji (5 g) was soaked in 25 mL of 10 mM acetic acid buffer (pH 5.0) containing 5% NaCl (w/v) for 3 h at room
temperature and then filtrated to obtain an enzyme extract solution. The a-amylase activity was measured using a
standard method 22, while a-glucosidase activity was measured using a glucoamylase and α-glucosidase assay kit
(Kikkoman Biochemifa Co., Ltd., Tokyo, Japan). One unit of a-glucosidase activity was defined as the titer that
100
101
102
103
104
105
106
107
released 1 mmol of p-nitrophenol from p-nitrophenyl-a-glucopyranoside in 1 min 25
.
Oligosaccharide and glycoside analysis
To detect oligosaccharides and glycosides, high-performance liquid chromatography (HPLC) was performed using
a SIL-20A direct injection autosampler, SCL-10Avp controller, and LC-20AD pump (Shimadzu Corporation, Kyoto,
Japan), with a Shodex SUGAR SZ5532 column (6.0 × 150 mm) (Showa Denko K.K., Tokyo, Japan). The column
temperature was maintained at 80°C using a forced-air circulation-type column oven (CTO-20A; Shimadzu
5
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 6 of 25
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
Corporation). A 5-µL aliquot of the sample was injected into the HPLC system. The flow rate was 0.9 mL/min and
the isocratic conditions were 25% Milli-Q water/75% acetonitrile. The system was also equipped with a Shodex RI-
501 refractive index detector (Showa Denko K.K.). For detection by mass spectrometry (MS), the ACQUITY UPLC
H-Class System (Waters Corporation, Milford, MA, USA) and XevoG2-XS high-performance benchtop quadrupole
time-of-flight (QTofTM) mass spectrometer (Waters Corporation) were used as described previously 26. When the
Shodex SUGAR SZ5532 column was used, 75% acetonitrile was used as the mobile phase at a flow rate of 0.9
mL/min. Quantification of oligosaccharides and glycosides were performed on HPLC-RI based on the standard
curves generated for each compound. For a-EM and a-EiM, compounds purified in this study were used to
generate the standard curves.
Protein purification
The EAGN strain was inoculated into 100 mL of YPM medium (yeast extract, peptone, and mannitol at 1%, 1% and
5%) in a 300-mL Erlenmeyer flask, cultured at 30°C for 6 days with shaking at 130 rpm, and then filtered through
Miracloth (Merck KGaA, Darmstadt, Germany) to remove the mycelium. The culture was filtered with a 50-kDa cut-
off ultrafiltration device (Biomax 50 kDa Ultrafiltration Discs; Merck KGaA) and a 100-kDa cut-off ultrafiltration
device (Amicon Ultra–15 mL Centrifugal Filters Ultracel-100K; Merck KGaA). It was then subjected to fast protein
liquid chromatography using the ÄKTA start chromatography system (GE Healthcare Japan Co., Tokyo, Japan),
equipped with a cation‐exchange column (HitrapTMSP Sepharose Fast Flow; GE Healthcare Japan Co.). Total
126
127
128
129
130
131
132
133
134
135
136
137
138
139
protein was measured using the method described by Lowry 27. Proteins in the samples were separated by sodium
dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) 28, on 12.5% polyacrylamide gels, and stained with
Coomassie brilliant blue R-250. The BlueStar Plus Prestained Protein Marker (MWP03; Nippon Genetics Europe
GmbH, Dueren, Germany) was used as a protein standard.
In vitro synthesis and glycoside purification
Purified AgdA (0.18 U/mL) was incubated with 40% (v/v) ethanol and 5% (w/v) maltose in 10 mM acetate buffer (pH
5.0) at 37°C for 48 h. The reaction was stopped by heating at 80°C for 10 min. Residual glucose in the reaction
mixture was removed by treating with S. cerevisiae K-901 cells. The cells were cultured in a three-fold volume of
GYP medium (glucose, yeast extract, and peptone at 2%, 0.5% and 0.5%) of in vitro reaction mixture at 30°C and
180 rpm for 24 h, and then centrifuged for 10 min at 900 x g. Afterward, the cells were rinsed with distilled sterile
water and then centrifuged twice to remove the medium components. Before adding the yeast, ethanol was
removed under reduced pressure and the in vitro reaction mixture was diluted with distilled water to adjust the total
sugar concentration to 5%. After adding yeast, the solution was incubated at 30°C and 180 rpm for 5 h. Thereafter,
6
ACS Paragon Plus Environment

Page 7 of 25
Journal of Agricultural and Food Chemistry
140
the yeast was removed by centrifugation followed by filtration using a disposable membrane filter unit with a pore
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
size of 0.45 mm (13HP045CN; Toyo Roshi Kaisha, Ltd., Tokyo, Japan). To purify the targeted compounds, yeast-
treated reaction solution was fractionated by HPLC with two columns: an Asahipak NH2P-130G 7B (7.5 × 50 mm)
(Showa Denko K.K.) and Asahipak NH2P-90 20F (20 × 300 mm) (Showa Denko K.K.). The introduction volume
was 100 mL and the flow rate was 5.0 mL/min. The isocratic conditions were 25% Milli-Q water/75% acetonitrile.
The collected fractions were concentrated by distillation under reduced pressure and separated by HPLC with the
use of 65% acetonitrile as the mobile phase to isolate the target compounds A and B.
Structural analysis of target compounds by nuclear magnetic resonance spectroscopy
Each saccharide samples was dissolved in 0.5 mL D2O (99.8 %), and exchangeable protons were substituted with
deuteriums by lyophilization. 1H, 13C, and two-dimensional (2D) NMR spectra were recorded using a RESONANCE
ECZ 600R nuclear magnetic resonance spectrometer (JEOL Ltd., Tokyo, Japan) at 14.1 T with D2O (99.8%) as the
solvent. To analyze compound A, residual 1H signals in D2O (known data: 3.34 ppm) 29 or 13C signals (known data:
69.7 ppm) 29 were used as an internal standard. To analyze compound B, acetone (dH: 2.225 ppm, dC:30.5 ppm)
was used as an internal standard. Signal assignments were based on unambiguous chemical shifts and
heteronuclear multiple bond coherence (C–H HMBC), heteronuclear single quantum coherence (C–H HSQC), and
correlation spectroscopy (H–H COSY).
Content of novel compounds in commercial sake
Following types of 111 commercial sakes were used: jummai-daiginjo (8 samples), jummai-ginjo (33 samples),
jummai (58 samples), daiginjo (1 sample), honjozo (4 samples), regular sake (4 samples), aged sake (1 sample),
kijo-shu (2 samples). Sake samples were filtered through a disposable membrane filter unit with a pore size of 0.45
mm, and analyzed by HPLC under the same conditions described in the section titled “oligosaccharide and
glycoside analysis.” For statistical analysis, student’s t-test was performed on Microsoft Excel 2016.
RESULTS AND DISCUSSION
Production of the control and DagdA strains
Oligosaccharides and glucosides formed by transglycosylation are present in sake. However, the enzymes
responsible for forming these compounds remain unclear. During the sake-brewing process, rice-koji made of rice
and filamentous fungus A. oryzae, provides enzymes in mash (moromi). AgdA is the main glucosidase produced by
A. oryzae; thus, AgdA is most likely involved in transglycosylation during the sake-brewing process. Therefore, the
transglycosylation products in sake were identified by comparing the compounds in sake brewed with rice-koji
7
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 8 of 25
172
173
174
175
176
177
178
179
180
made of A. oryzae DagdA and control strains. A DNA fragment for DagdA was constructed and transformed into the
A. oryzae recipient strain NS4DligD. Among several candidate transformed strains, one was confirmed by Southern
blot and PCR analyses as an DagdA strain (Figures S1A and S1B, respectively). To eliminate the auxotrophic effect
of the DagdA strain on rice growth, a control strain, named NSCS, was also constructed by transformation of strain
NS4DligD with the selective marker gene sC (Figure S2). Both these strains were used for rice-koji production. The
hydrolytic enzyme activities and amounts of mycelia in the resulting rice-koji were analyzed. While the hydrolytic
enzyme activities and amount of mycelia were similar between the DagdA and NSCS strains, a-glucosidase activity
of the DagdA strain decreased significantly (p<0.05) by about 50% of that of the NSCS strain, indicating AgdA is
responsible for at least half of the a-glucosidase activity in rice-koji (Figure S3). Although AgdA is considered the
main glucosidase, this is the first report to demonstrate AgdA’s contribution to a-glucosidase activity in the sake-
brewing process.
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
Analysis of oligosaccharides in sake brewed using the DagdA strain
To investigate AgdA’s role in forming transglycosylation compounds in sake, sake was brewed on a small scale
using the rice-koji of the DagdA and NSCS strains. Curves of cumulative weight loss, which reflect the loss of CO2
gas accompanied with alcoholic fermentation, were similar in both fermentations, but the sake meter value and the
alcohol content were slightly higher with the use of the DagdA strain (Table S2). The increased alcohol level
observed in the sample of the DagdA strain may be due to reduced formation of unfermentable sugars by
transglycosylation. However, hydrophilic interaction liquid chromatography/TOF-MS analysis showed increases in
the isomaltose and kojibiose levels from 800 mg/L and 270 mg/L to 2,409 mg/L and 692 mg/L, respectively, in the
sake brewed with the DagdA strain (Figure 1A), suggesting that AgdA plays a role in degrading isomaltose and
kojibiose during the sake-brewing process. Because AgdA is the main a-glucosidase produced by A. oryzae, it was
expected to be one of the factor(s) responsible for formation of isomaltose, the most abundant oligosaccharide.
However, the sake brewed with rice-koji of A. oryzae strain DagdA contained larger amounts of isomaltose
196
197
198
199
200
201
indicated that AgdA’s contribution for isomaltose formation is not important. Notably, a-EG and a-GG levels in the
sake brewed with the DagdA strain were markedly decreased from 3,180 mg/L and 2954 mg/L to 824 mg/L and
1389 mg/L, respectively, (Figure 1A), which clearly indicates that AgdA is responsible for glycoside formation during
the sake-brewing process. This is the first report to identify the enzyme responsible for a-EG and a-GG formation
during the sake-brewing process. Other than those, a peak observed at 1.64 min (m/z=213.07) was increased in
the sake brewed with the DagdA strain, which we could not identify so far.
8
ACS Paragon Plus Environment

Page 9 of 25
Journal of Agricultural and Food Chemistry
202
Interestingly, two peaks were observed at 4.4 min and 5.2 min, which we inferred as compounds A and B.
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
However, the amounts of compounds A and B decreased significantly in the sake brewed with the DagdA strain
(Figure 1A). TOF/MS analysis showed the exact masses of lithium adduct ions ([M+Li]+) of both compounds were
377.16 (Figure 1B), suggesting the compounds were structural isomers. The m/z value of 377.16 coincides with
those of the dehydration and condensation products of maltose and ethanol. This finding is consistent with the
observation that the loss of AgdA, which likely catalyzes the dehydration and condensation of maltose and ethanol
during the sake-brewing process, resulted in decreased amounts of compounds A and B. Comparisons of MS/MS
fragmentation peaks of a-EG with those of compounds A and B showed several overlapping peaks (Figure 1C).
The m/z values of the fragmented peaks generated from compounds A and B were almost identical, but with
different intensities. Reportedly, the MS/MS fragment intensities differ between structural isomeric saccharides,
supporting the idea that compounds A and B are structural isomers. Detailed assignment of the fragment of a-EG
+
+
indicated that overlapped fragments 0,3A1+ or/and 1,4A1 , 0,2A1 , and B1+ are derived from the nonreductive ends of
glucose residues (Figure 1D). Taken together, these findings strongly suggest that compounds A and B are glycosyl
products of a-EG; namely, ethyl a-maltoside. In addition, compounds with the exact masses of the lithium adduct
were identical to ethyl maltotorioside and maltotetraoside (m/z = 539.22 and 701.27, respectively), suggesting that
transglycosylation by AgdA contributes to sake’s component complexity.
Production of an AgdA by agdA overexpression strain
To identify the structures, compounds A and B were purified from sake to obtain sufficient amounts for NMR
analysis. However, due to sake’s complex components, it was impossible to purify these compounds directly.
Therefore, in vitro synthesis by enzymatic reactions was attempted. To prepare AgdA, an AgdA overexpression
strain of A. oryzae was constructed, where agdA was expressed under a strong promoter containing multiple
copies of cis-elements for amylase induction17. A. oryzae strain niaD300 was transformed with the overexpression
vector and a strain harboring two copies of the expression vector, named as EAGN, was selected by Southern blot
analysis (Figures S4A and B). EAGN was cultured in YPM liquid medium for 6 days, and a-glucosidase activity of
227
228
229
230
231
232
the culture filtrates reached 0.298 U/mL, which was 75-fold greater than the a-glucosidase activity of the wild-type
A. oryzae strain RIB40 (Table 1). AgdA was prepared by ultrafiltration and fast protein liquid chromatography, which
finally yielded a purified AgdA protein (Table 1). SDS-PAGE analysis showed a single band corresponding to the
size of AgdA (ca. 109 kDa)14 in lane B, while several signals were observed in the culture filtrate of EAGN in lane A,
suggesting that AgdA was successfully purified (Figure 2).
Enzymatic synthesis and purification of the target compounds
9
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 10 of 25
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
Compounds A and B were synthesized in vitro, using ethanol and maltose as substrates, and purified AgdA at
concentrations of 20%–60% (v/v), 1%–10% (w/v) and 0.01–0.18 U/mL, respectively, to determine the optimal
reaction conditions. Consequently, the optimal substrate and enzyme concentrations for synthesizing the target
compounds were 40% ethanol, 5% maltose, and 0.18 U/mL of purified AgdA. HPLC analysis of the products of in
vitro reaction showed several peaks (Figure 3B). There were obvious peaks of maltose, glucose, and a-EG on a
chromatogram. In addition, there were peaks of the in vitro reaction solution with elution times corresponding to
those of compounds A and B in sake (Figures 3A and B). LC-qTOF/MS confirmed that the m/z values ([M+Li]+) of
the compounds corresponding to the peaks were 337.16, and MS/MS fragmentation peaks were identical to the
fragment peaks of compounds A and B (Figures 4A and B), indicating that the compounds synthesized in vitro were
identical to compounds A and B. Subsequently, attempts were made to purify compounds A and B; however, there
was a large peak corresponding to glucose, as a by-product, which was eluted immediately after compound B, but
prevented fractionation of compound B. Because yeast readily consumes glucose in medium, it was added to
eliminate glucose from the reaction solution for 5 h at 30°C. As a result, glucose was removed successfully from
the solution below the detection limit, and a peak corresponding to compound B was quite distinct (Figure 3C). On
the other hand, along with glucose consumption, glycerol was produced in the solution, although the peak was
overlapped with that of compound A. Nonetheless, glycerol and compound A were separated with the use of a
preparative column (Asahipak NH2P-130G 7B and Asahipak NH2P-90 20F). Finally, compounds A and B were
fractionated and lyophilized. Figures 3D and E show chromatograms of the purified fractions of compounds A and
B, respectively. The amounts of purified compounds A and B were 10.6 and 128.1 mg, respectively.
Structural determination by NMR
The purified compounds were subjected to determine the structures. Each compound showed two glycosyl
residues and one ethoxy segment. 1H and 13C NMR data of compound A (Table S3, Figure S5) were identical to
those of the synthetic ethyl a-maltoside (Figure 5A) 29. While no NMR data associated with compound B was
previously reported, further analysis including DQF-COSY, HSQC, HSQC-TOCSY, and HMBC was performed.
Analysis was initiated from two anomeric protons at 4.94 and 4.95 ppm (Table S4, Figure S6). After confirmation of
C-H correlation using HSQC, each signal was classified into three residues, two saccharide residues, and one ethyl
group by HSQC-TOCSY (Figure S7B, C). Spin networks in each residue were confirmed mainly by DQF-COSY
(Figure S7A). In both glycosyl residues, proton that showed correlation peak with anomeric proton was assigned as
2-H. Another proton showed correlation with 2-H was 3-H, and the other coupling partner of 3-H was assigned as
4-H. 5-H showed correlation peaks with 4-H and two 6-H. 3JH-H coupling constants between 2-H/3-H, 3-H/4-H, and
4-H/5-H were all in range of 9.0–10.5 Hz. This finding suggested that these protons have axial orientation and the
10
ACS Paragon Plus Environment

Page 11 of 25
Journal of Agricultural and Food Chemistry
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
saccharide residues are glucopyranosyl. Pyranose form was also confirmed by a HMBC cross peak between H-
1/C5 (Figure S7D). a-Glycosidic linkages were suggested by gauche (3.9 Hz) coupling constants of anomeric
protons. Inter-residual orientation was determined by HMBC correlation peaks between one of the anomeric
protons (1-H: 4.94 ppm) and carbon of ethyl group (2”C: 64.4 ppm), and the C6 carbon (C6: 65.8 ppm) and 1-H (1’:
4.95 ppm) of the other glucosyl residue (Figure 5C). As a result, the structure of compound B was determined to be
ethyl a-isomaltoside (Figure 5B) 30. This is the first report of these compounds in sake. Organic synthesis of a-EM
29 and enzymatic synthesis of a-EiM 30 were reported previously, but this is the first study to report that both are
natural products. In addition, the results of the present study also demonstrated that transglycosylation activity of
AgdA synthesizes these glycosides during the sake-brewing process because the amounts of compounds A and B
were low in sake brewed with A. oryzae, lacking a functional AgdA, while both were synthesized successfully in
vitro with purified AgdA. It is likely that AgdA transfers the glucose residue at non-reducing end of maltose to 4-OH
or 6-OH of the glucose residue in a-EG.
Content of novel compounds in commercial sake
The amounts of these novel compounds in commercial sake products were determined. We determined
concentrations of a-EM and a-EiM in 111 sake samples. The amounts of them varies among sakes but they are
commonly contained in sake (Figure 6A, B, Table 2), The average concentrations of a-EM and a-EiM were 100 and
498 mg/L, respectively. Comparing the average concentrations of glucose and isomaltose (2.22% and 0.31%),
which are appropriate levels as sake 1, 3, the concentration of a-EM and a-EiM were one-tenth to one-hundredth of
those sugars. However, it is notable that several sakes contained a-EiM at over 1,000 ppm. Although the sensory
properties of both a-EM and a-EiM have not been clarified in this study, these glycosides may affect the taste of
sake.
285
286
287
288
289
290
291
292
293
294
295
The amount of a-EiM was five-fold greater than a-EM; there was a similar tendency in the in vitro reaction by
AgdA (Figure 3B). These results probably reflect transglycosylation’s preference for the 6-OH of the glucose rather
than the 4-OH of a-EG, and/or the preference of hydroxylation of a-EM over a-EiM. There was weak correlation
between the concentrations of a-EM and a-EG (r=0.63), while no correlation was observed between a-EiM and a-
EG (r=-0.10), and a-EM and a-EiM (r=0.19), indicating that formation of a-EM and a-EiM did not simply depend on
the amounts of the acceptor substrate. Comparison of the average concentrations of the a-EM, a-EiM and a-EG
among sake types is shown in Table 2. Statistical analysis showed that the averages concentration of a-EM
between jummai and jummai-daiginjo, are significantly different (p<0.05), but other than that, we could not find any
significant differences in the concentration of the glycosides among sake types. Future studies using sake samples
11
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 12 of 25
296
297
with more detailed information on production process are highly desirable to identify the factor(s) involved in
forming novel glycosides.
12
ACS Paragon Plus Environment

Page 13 of 25
Journal of Agricultural and Food Chemistry
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
Abbreviations Used
HPLC, High-performance liquid chromatography; MS, Mass spectrometry; ORF, Open reading frame; PCR,
Polymerase chain reaction; DQF-COSY, double quantum filter-correlation spectroscopy; HSQC, heteronuclear
single quantum coherence; HSQC-TOCSY, heteronuclear single quantum coherence-Total correlation
spectroscopy; HMBC, heteronuclear multiple bond coherence
Acknowledgment
We wish to thank Katsuya Gomi (Tohoku University) for donating the fungal host strain NS4DligD. We also wish to
thank Akio Koda (Ozeki Co., Ltd.) for generously providing the high expression vector for A. oryzae.
Funding
This work was supported by a grant from JSPS KAKENHI (no. JP17K15274).
Supporting Information
Figure S1, Construction of strain ΔagdA.; Figure S2, Confirmation of the control strain, NSCS, by PCR analysis.;
Figure S 3, Activities of a-amylase and a-glucosidase of the a-agdA and control (NSCS) strains.; Figure S4,
Confirmation of introduction of the AgdA overexpression cassette.; Figure S5, 1H NMR and 13C NMR spectra of
compound A in D2O.; Figure S6, 1H NMR and 13C NMR spectra of compound B in D2O.; Figure S7, 2D NMR data of
compound B.; Table S1, Primers used for PCR in this study.; Table S2, Analytical values of the sakes produced
using the rice-koji of DagdA and NSCS strains.; Table S3, NMR analysis of compound A.; Table S4, NMR analysis
319
320
321
of compound B. Table S5, Average concentrations of the a-EM, a-EiM and a-EG among sake types.
13
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 14 of 25
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
References
1. Fujii, T.; Isogai, A.; Izu, H.; Kanda, R.; Kizaki, Y. Analysis of sake components presented to sake contest in 2014.
Rep. Res. Inst. Brew. 2015, 187, 1–16.
2. Iwatsuki, M. Studies on Glycerin in Sake-Brewing. J. Brew. Soc. Japan. 1962, 57, 329–332.
3. Aso, K.; Shibasaki, K.; Yamauchi, F. Studies on the Unfermentable Sugars (VII). J. Ferment. Technol. 1954, 32,
47–52
4. Blomberg, A. Metabolic surprises in Saccharomyces cerevisiae during adaptation to saline conditions. FEMS
Microbiol. Lett. 2000, 182, 1–8.
5. Baba, S.; Oguri, I.; Fukuzawa, M.; Iida. T.; Kobayashi, I.; Imai, K. Kojikinkouso niyoru mushimai kara origotou no
seisei ni tuite (in Japanese). J. Brew. Soc. Japan. 1974, 69, 781–783.
6. Oka, S.; Sato, S. Contribution of ethyl α-D-glucoside to flavor construction in sake. J. Agric. Chem. Soc. Japan.
1976, 50, 455–461.
7. Takenaka, F.; Uchiyama, H.; Imamura, T. Identification of α-D-glucosylglycerol in sake. Biosci. Biotechnol.
Biochem. 2000, 64, 378–385.
8. Kitamura, N.; Ota, Y.; Haratake, A; Ikemoto, T.; Tanno, O.; Horikoshi, T. Effects of ethyl alpha-D-glucoside on skin
barrier disruption. Skin Pharmacol. 1997, 10, 153–159.
9. Hirotsune, M.; Haratake, A.; Komiya, A.; Sugita, J.; Tachihara, T.; Komai, T.; Hizume, K.; Ozeki, K.; Ikemoto, T.
Effect of ingested concentrate and components of sake on epidermal permeability barrier disruption by UVB
irradiation. J. Agric. Food Chem. 2005, 53, 948–952.
10. Hirotsune, M.; Haratake, A.; Komiya, A.; Sugita, J.; Tachihara, T.; Komai, T.; Hizume, K.; Ozeki, K.; Ikemoto, T.
Effects of topical application of alpha-D-glucosylglycerol on dermal levels of insulin-like growth factor-i in mice and
on facial skin elasticity in humans. Biosci. Biotechnol. Biochem. 2010, 74, 759–765.
11. Oka, S.; Iwano, K.; Nunokawa, Y. Formation of ethyl a-D-glucoside in sake brewing. J. Agric. Chem. Soc.
Japan. 1976, 50, 463–468.
12. Takenaka, F.; Uchiyama, H. Synthesis of alpha-D-glucosylglycerol by alpha-glucosidase and some of its
characteristics. Biosci. Biotechnol. Biochem. 2000, 64, 1821–1826.
13. Pazur, J.; French, D. The action of transglucosidase of Aspergillus oryzae on maltose. J. Biol. Chem. 1952,
196, 265–272.
14. Minetoki, T.; Gomi, K.; Kitamoto, K.; Kumagai, C.; Tamura, G. Nucleotide sequence and expression of alpha-
glucosidase-encoding gene (agdA) from Aspergillus oryzae. Biosci. Biotechnol. Biochem. 1995, 59, 516–521.
14
ACS Paragon Plus Environment

Page 15 of 25
Journal of Agricultural and Food Chemistry
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
15. Minetoki, T.; Gomi, K.; Kitamoto, K.; Kumagai, C.; Tamura, G. Characteristic expression of three amylase-
encoding genes, agdA, amyB, and glaA in Aspergillus oryzae transformants containing multiple copies of the agdA
gene. Biosci. Biotechnol. Biochem. 1995, 59, 2251-2254.
16. Yamada, O.; Lee, B.; Gomi, K. Transformation system for Aspergillus oryzae with double auxotrophic
mutations, niaD and sC. Biosci. Biotechnol. Biochem. 1997, 61, 1367–1369.
17. Minetoki, T.; Kumagai, C.; Gomi, K.; Kitamoto, K.; Takahashi, K. Improvement of promoter activity by the
introduction of multiple copies of the conserved region III sequence, involved in the efficient expression of
Aspergillus oryzae amylase-encoding genes. Appl. Microbiol. Biotechnol. 1998, 50, 459–467.
18. Mizutani, O.; Kudo, Y.; Saito, A.; Matsuura, T.; Inoue, H.; Abe, K.; Gomi, K. A defect of LigD (human Lig4
homolog) for nonhomologous end joining significantly improves efficiency of gene-targeting in Aspergillus oryzae.
Fungal. Genet. Biol. 2008, 45, 878–889.
19. Gomi, K.; Iimura, Y.; Hara, S. Integrative transformation of Aspergillus oryzae with a plasmid containing the
Aspergillus nidulans argB gene. Agric. Biol. Chem. 1987, 51, 2549–2555.
20. Arakawa, GY.; Kudo, H.; Yanase, A.; Eguchi, Y.; Kodama, H.; Ogawa, M.; Koyama, Y.; Shindo, H.; Hosaka, M.;
Tokuoka, M. A unique Zn(II)2-Cys6-type protein, KpeA, is involved in secondary metabolism and conidiation in
Aspergillus oryzae. Fungal Genet. Biol. 2019, 127, 35–44.
21. Tokuoka, M.: Sawamura, N.; Kobayashi, K.; Mizuno, A. Simple metabolite extraction method for metabolic
profiling of the solid-state fermentation of Aspergillus oryzae. J. Biosc. Bioeng. 2010, 110, 665–669.
22. The Brewing Society of Japan. Official methods of analysis of National Tax Administration Agency, 4th ed. The
Brewing Society of Japan, Tokyo, Japan, 1993, 13–221.
23. Reissig, JL.; Storminger, JL.; Leloir, LF. A modified colorimetric method for the estimation of N-acetylamino
sugars. J. Biol. Chem. 1955, 217, 959–966.
24. Fujii, F.; Ozeki, K.; Kanda, A.; Hamachi, M.; Nunokawa, Y. A simple method for the determination of grown
mycelial content in rice-koji using commercial cell wall lytic enzyme, Yatalase. J. Brew. Soc. Japan. 1992, 87, 757–
759.
25. Imai, Y.; Tokutake, S.; Yamaji, N.; Suzuki, M. A new method of measuring glucose-forming activity in rice koji. J.
Brew. Soc. Japan. 1996, 91, 51–57.
26. Tokuoka, M.; Honda, C.; Totsuka, A.; Shindo, H.; Hosaka, M. Analysis of the oligosaccharides in Japanese rice
wine, sake, by hydrophilic interaction liquid chromatography-time-of-flight/mass spectrometry. J. Biosci. Bioeng.
2017, 124, 171–177.
27. Lowry, OH.; Rosebrough, NJ.; Farr, AL.; Randall, RJ. Protein measurement with the Folin phenol reagent. J.
Biol. Chem. 1951, 193, 265–275.
15
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 16 of 25
384
385
386
387
388
389
390
391
392
393
28. Laemmli, UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature.
1970, 227, 680–685.
29. Tanaka, T.; Kikuta, N.; Kimura, Y.; Shoda, S. Metal-catalyzed Stereoselective and Protecting-group-free
Synthesis of 1,2-cis-Glycosides Using 4,6-Dimethoxy-1,3,5-triazin-2-yl Glycosides as Glycosyl Donors. Chem. Lett.
2015, 44, 846–848.
30. Kim, YK.; Tsumuraya, Y.; Sakano, Y. Enzymatic preparation of novel non-reducing oligosaccharides having an
isomaltosyl residue by using the transfer action of isomaltodextranase from Arthrobacter globiformis T6. Biosci.
Biotech. Biochem. 1995, 59, 1367–1369.
394
16
ACS Paragon Plus Environment

Page 17 of 25
Journal of Agricultural and Food Chemistry
395
396
397
398
399
Figure Legend
Figure 1. Analysis of sake, brewed with strain DagdA by hydrophilic interaction liquid chromatography/TOF-MS. (A)
MS chromatogram (TIC) of sake, brewed with the DagdA and control NSCS strains.1, a-EG; 2, compound A; 3,
compound B; 4,a-GG; 5, sakebiose; 6, kojibiose; and 7, isomaltose. The peak intensities of compounds A and B in
the sake brewed with strain DagdA were reduced significantly. (B) MS spectrum of compound A (RT; 4.430 min) and
B (RT; 5.172 min). (C) MS/MS fragmentation peaks of compound A, B, and a-EG. (D) Detailed assignment of the
MS/MS fragment of a-EG.
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
Figure 2. SDS-PAGE of culture filtrates of strain EAGN (day 6) (lane a) and purified AgdA (lane b) after ultrafiltration
and fast protein liquid chromatography.
Figure 3. Chromatogram of (A) commercial sake, (B) in vitro reaction mixture, (C) yeast-treated reaction solution,
(D) purified compound A and (E) purified compound B by HPLC-RI. 1, a-EG; 2, compound A; 3, compound B; 4, a-
GG; 5, sakebiose; 6, kojibiose; 7, isomaltose; 8, glycerol; 9, glucose; and 10, maltose.
Figure 4. MS/MS analysis of sake components and synthetic products. (A) compound A (RT; 4.430 min), (B)
compound B (RT; 5.172 min).
Figure 5. Chemical structure of (A) compound A (ethyl a-maltoside) and (B) compound B (ethyl a-isomaltoside),
and (C) DQF-COSY and key HMBC correlations of compound B.
416
417
Figure 6. Histogram of the distribution of (A) ethyl a-maltoside and (B) ethyl a-isomaltoside content in 111
commercial sakes.
17
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 18 of 25
Tables
Table 1. a-Glucosidase activities in the culture broths and purified AgdA
α-Glucosidase
Total protein
amount
Specific
activity
Purification
Volume
Total activity
activity
(U/ml)
degree
(fold)
(U/mg
protein)
(µL)
(U)
(mg)
Culture broth (NGAG)
Purified AgdA
290.0
53.4
-
0.298
0.230
0.004
86.28
12.08
-
1662.17
0.052
5.631
-
1.0
108.5
-
2.15
-
Culture broth (niaD300)
Table 2. Average concentrations of the a-EM, a-EiM and a-EG among sake types.
a-EM
a-EiM
a-EG
Number of
samples
mg/L
jummai-daiginjo
jummai-ginjo
jummai
58
102
111
31
503
563
467
389
515
422
316
581
3,675
4,954
5,565
4,366
2,784
4,010
2,568
3,323
8
33
58
1
daiginjo
honjozo
84
4
regular sake
aged sake
kijo-shu
60
4
34
1
79
2
Average
100
498
5,012
18
ACS Paragon Plus Environment

Page 19 of 25
Figures
Journal of Agricultural and Food Chemistry
19
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 20 of 25
20
ACS Paragon Plus Environment

Page 21 of 25
Journal of Agricultural and Food Chemistry
21
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 22 of 25
22
ACS Paragon Plus Environment

Page 23 of 25
Journal of Agricultural and Food Chemistry
23
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 24 of 25
24
ACS Paragon Plus Environment

Page 25 of 25
Journal of Agricultural and Food Chemistry
For Table of Contents Only
89x50mm (96 x 96 DPI)
ACS Paragon Plus Environment