Enhanced Synthesis of 2-O-α- d -Glucopyranosyl- l -ascorbic Acid from α-Cyclodextrin by a Highly Disproportionating CGTase

Article abstract of DOI:10.1021/acscatal.5b02108

2-O-α-d-Glucopyranosyl-l-ascorbic acid (AA-2G) is an industrially important derivative of vitamin C [l-ascorbic acid (l-AA)]. A useful synthetic route toward AA-2G is the selective glucosylation of l-AA by cyclodextrin glucanotransferase (CGTase). However, the cyclodextrin donor substrate is utilized rather inefficiently, because only one of its constituent glucosyl residues is coupled to the l-AA acceptor. A CGTase catalyzing disproportionation of the linear maltooligosaccharide chain formed in the initial coupling reaction might utilize a greater portion of the substrate for l-AA glucosylation and thus boost the AA-2G yield of cyclodextrin conversion. We present here a detailed characterization of the transfer reactions involved in the formation of AA-2G from α-cyclodextrin by a commercial CGTase preparation from Thermoanaerobacter sp. (Toruzyme 3.0L). We demonstrate that besides coupling, disproportionation constitutes a major route of glucosylation of l-AA by this enzyme. l-AA glucosides with oligoglucosyl chains between 1 and 12 units long were produced in the reaction. After chain-trimming hydrolysis with glucoamylase, however, AA-2G was recovered as the sole product of the enzymatic transglucosylation. The molar yield of AA-2G from cyclodextrin was 1.4, thus clearly exceeding the maximal yield of 1 for the coupling reaction. Using conditions optimized for transfer efficiency and productivity, we obtained AA-2G at the highest concentration (143 g/L, 424 mM) so far reported from an enzymatic glucosylation of l-AA. The synthetic yield was 30% based on l-AA (250 g/L, 1420 mM) offered in ≤4.6-fold molar excess over α-cyclodextrin.

Full text of DOI:10.1021/acscatal.5b02108

Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI
Article
Enhanced synthesis of 2-O-#-D-glucopyranosyl-L-ascorbic
acid from #-cyclodextrin by a high disproportionating CGTase
Rama Krishna Gudiminchi, Andrew Towns, Subhash Varalwar, and Bernd Nidetzky
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.5b02108 • Publication Date (Web): 22 Jan 2016
Downloaded from http://pubs.acs.org on January 25, 2016
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 free 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 accessible to all
readers and 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.
ACS Catalysis 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 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
Enhanced synthesis of 2ꢀOꢀαꢀ
D
ꢀglucopyranosylꢀ ꢀ
L
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ascorbic acid from αꢀcyclodextrin by a high
disproportionating CGTase
Rama Krishna Gudiminchi†, Andrew Towns‡, Subhash Varalwar¶, Bernd Nidetzky*†§
†Austrian Center of Industrial Biotechnology, Petersgasse 14, Aꢀ8010 Graz, Austria
‡Vivimed Labs Europe Ltd, PO Box B3, Leeds Road, Huddersfield, West Yorkshire, HD1 6BU,
United Kingdom
¶Vivimed Labs Ltd, Veeranag Towers, Habsiguda, Hyderabadꢀ500007, Telangana, India
§Institute of Biotechnology and Biochemical Engineering, Graz University of Technology, 12/1
Petersgasse, Aꢀ8010 Graz, Austria
ABSTRACT: 2ꢀOꢀαꢀ
derivative of vitamin C (
selective glucosylation of
cyclodextrin donor substrate is utilized rather inefficiently, because only one of its constituent
glucosyl residues is coupled to the ꢀAA acceptor. A CGTase catalyzing disproportionation of
D
ꢀGlucopyranosylꢀ
ꢀascorbic acid;
ꢀAA by cyclodextrin glucanotransferase (CGTase). However, the
L
ꢀascorbic acid (AAꢀ2G) is an industrially important
L
LꢀAA). A useful synthetic route towards AAꢀ2G is the
L
L
the linear maltooligosaccharide chain formed in the initial coupling reaction might utilize a
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 37
1
2
3
4
greater portion of the substrate for
L
ꢀAA glucosylation and thus boost the AAꢀ2G yield on
5
6
7
8
cyclodextrin conversion. We present here a detailed characterization of the transfer reactions
involved in AAꢀ2G formation from αꢀcyclodextrin by a commercial CGTase preparation from
Thermoanaerobacter sp. (Toruzyme 3.0L). We demonstrate that besides coupling,
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
disproportionation constitutes a major route of glucosylation of LꢀAA by this enzyme. LꢀAA
glucosides with oligoglucosyl chains between 1 and 12 units long were produced in the reaction.
After chainꢀtrimming hydrolysis with glucoamylase, however, AAꢀ2G was recovered as sole
product of the enzymatic transglucosylation. The molar yield of AAꢀ2G from cyclodextrin was
1.4, thus clearly exceeding the maximum yield of 1 for the coupling reaction. Using conditions
optimized for transfer efficiency and productivity, we obtained AAꢀ2G in the highest
concentration (143 g/L; 424 mM) so far reported from an enzymatic glucosylation of
LꢀAA. The
synthetic yield was 30% based on
LꢀAA (250 g/L; 1420 mM) offered in ≤ 4.6ꢀfold molar excess
over αꢀcyclodextrin.
Key words: Ascorbic acid glucoside, AA-2G production, CGTase, Thermoanaerobacter sp.,
disproportionation, α-cyclodextrin, starch
INTRODUCTION
Biocatalytic transglycosylations (Scheme 1) constitute important synthetic routes towards
glycosides as industrial chemicals.1ꢀ4 Their optimization typically requires a carefully integrated
development, where the enzyme used and the reaction scheme applied are strongly intertwined
key factors of overall process performance.5ꢀ9 Here we focus on AAꢀ2G synthesis through
CGTaseꢀcatalyzed transglucosylation from αꢀcyclodextrin to LꢀAA (Figure 1).
ACS Paragon Plus Environment
2

Page 3 of 37
ACS Catalysis
1
2
3
4
LꢀAA is an important commodity chemical having wide applications with main uses in foods
5
6
7
8
and dietary supplements.10ꢀ13 However,
scope of its applications.14 Modification at the 2ꢀOH (e.g. glycosylation,14 phosphorylation15)
LꢀAA is highly prone to degradation, thus limiting the
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
effectively stabilizes the LꢀAA. AAꢀ2G has drawn high interest for application in the cosmetic,
food and healthꢀcare areas.16 Currently, AAꢀ2G is mainly used as an active ingredient in skinꢀ
care products.
In order to form the αꢀglucoside of ꢀAA by transglucosylation (Scheme 1), an αꢀconfigured
L
glucoside must be used as the donor substrate of the enzymatic reaction. Starchꢀderived maltoꢀ
oligosaccharides are therefore promising in particular. Not only do they constitute a group of
relatively inexpensive αꢀglucosidic substrates, but there are also industrial enzymes available as
possibly useful,
L
ꢀAA glucosylating catalysts.17 High siteꢀselectivity of the enzyme used is
however essential to prevent glucosylation of
L
ꢀAA at positions other than the 2ꢀOH (Scheme
2).18–20 Furthermore, enzymes catalyzing transglucosylation normally also catalyze, to a widelyꢀ
varying degree, the reaction with water (hydrolysis; Scheme 1).21 The class of glycoside
hydrolases reflects a continuum of enzyme selectivities between the (scarcely hydrolyzing)
transglycosidases and the true hydrolases.22,23 An enzyme low in αꢀglucoside hydrolase activity
is therefore needed to avoid "unproductive" utilization of donor substrate during synthesis of
AAꢀ2G. Among different options considered, including αꢀglucosidase and disproportionating 4ꢀ
αꢀglucanotransferase, CGTase was a promising candidate catalyst.18,21,24,25
CGTase (EC 2.4.1.19) is a special 4ꢀαꢀglucanotransferase widely known for its application to
the industrial synthesis of cyclodextrins from starch.26 Besides cyclization, CGTase also
catalyzes coupling, disproportionation, and hydrolysis reactions, as shown in Scheme 1.27
Considering transglucosylation from cyclodextrin, both coupling and disproportionation are
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 37
1
2
3
4
5
6
7
8
9
synthetically relevant reactions. The known CGTases differ in the extent to which
disproportionation participates in the overall enzymatic reaction, the synthesis of alkyl
glucosides for example.28 Transglucosylation exclusively via coupling is limited in that just a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
single glucosyl residue is attached to the
L
ꢀAA from the cyclic oligoꢀglucosyl donor substrate.
The product formed initially is thus a 2ꢀOꢀαꢀoligoꢀglucosyl derivative of
L
ꢀAA.19 Hydrolysis of
the oligoꢀglucosyl chain by either the endogenous exoꢀglucosidase activity of the CGTase29 or a
subsequently added glucosidase19 is used to recover the AAꢀ2G product. However, all nonꢀ
coupled glucosyl residues are consequently lost and cyclodextrin substrate is therefore utilized
inefficiently.29 Transglucosylation via disproportionation of oligoꢀglucosyl chains derived
through coupling or hydrolysis could thus enhance significantly the yield of AAꢀ2G on
cyclodextrin converted.
Previous studies of CGTaseꢀcatalyzed synthesis of AAꢀ2G focused on the coupling reaction
from cyclodextrin.29 Thermostable CGTases from Bacillus sp. were mainly used. The problem of
incomplete utilization of the glucosyl residues from the cyclodextrin substrate was recognized.
Protein engineering was applied to broaden the substrate scope of Paenibacillus macerans
CGTase to less expensive and, from the perspective of glucosyl content, more completely
utilizable substrates (e.g. maltose).30ꢀ32 It was noted that increased disproportionation activity of
enzyme variants could be favorable for the AAꢀ2G synthesis.32 However, the possible useful
involvement of disproportionation in AAꢀ2G synthesis was not examined, whereas the study we
report here characterizes its role in detail, specifically transglucosylation from αꢀcyclodextrin to
L
ꢀAA by a high disproportionating CGTase (from Thermoanaerobacter sp.; Toruzyme 3.0L).33,34
The evidence presented is relevant in showing significant (1.4ꢀfold) enhancement in AAꢀ2G
yield due to disproportionation, which is demonstrated to occur from intermediary oligoꢀ
ACS Paragon Plus Environment
4

Page 5 of 37
ACS Catalysis
1
2
3
4
glucosylated forms of
L
ꢀAA to free
LꢀAA. By exploiting this type of disproportionation, we
5
6
report enzymatic AAꢀ2G synthesis in the highest concentrations and yields so far disclosed.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
Materials and Assays: Toruzyme 3.0L, a commercial liquid preparation of CGTase from
Thermoanaerobacter sp., was kindly provided by Novozymes A/S (Bagsvaerd, Denmark).
Aspergillus niger glucoamylase, A. niger αꢀglucosidase and rice αꢀglucosidase were from
SigmaꢀAldrich (Schnelldorf, Germany). αꢀCyclodextrin,
LꢀAA, maltohexaose and AAꢀ2G
(>98%) were from TCI Deutschland GmbH (Eschborn, Germany). Dowex 1X2 (chloride form,
50ꢀ100 mesh) and calcined Celite 501 were from SigmaꢀAldrich (Vienna, Austria). Activated
charcoal was from Merck (Darmstadt, Germany). All other chemicals were from Carl Roth
GmbH+Co. KG (Karlsruhe, Germany) unless specified otherwise.
Protein concentration was determined using RotiꢀQuant assay (Roth) referenced against bovine
serum albumin. The Toruzyme preparation contained 5 mg/mL of protein content and its
activities were 114 U/mL and 161 U/mL of coupling and cyclization respectively as previously
determined.35,36 Toruzyme is a technicalꢀgrade enzyme preparation. Its enzymatically active
component is the CGTase from Thermoanaerobacter sp., as declared by the supplier and
demonstrated in several previous studies.28,35,37ꢀ39 The αꢀglucosidase and glucoamylase from A.
niger were commercially supplied as powder with enzyme activity of 892 U/g and 70,000 U/g
respectively. Stock solutions of A. niger αꢀglucosidase (134 U/mL) and glucoamylase (1000
U/mL) were prepared by dissolving 150 mg and 14.3 mg of αꢀglucosidase and glucoamylase
respectively in one mL of 100 mM sodium citrate buffer (pH 4.5). Commercial preparation of
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 37
1
2
3
4
rice αꢀglucosidase was supplied as ammonium sulfate suspension with activity and protein
5
content of 134 U/mL and 1.68 mg/mL respectively.
6
7
8
9
Reaction Set-up For Enzymatic Transglucosylation: Transglucosylation of
LꢀAA was
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
performed in 1.5 mL Eppendorf tubes using 100 mM sodium citrate buffer (pH 4.5), containing
2.6 mM thiourea as reducing agent. Using Toruzyme, αꢀcyclodextrin (up to 310 mM) or
maltohexaose (310 mM) was used as donor substrate. Using A. niger or rice αꢀglucosidase,
maltose (310 mM) was used as donor substrate. LꢀAA (up to 1420 mM) was used as acceptor
substrate. The ratio of acceptor to donor substrate was varied as part of the reaction optimization.
Temperature was also varied in the range 30 ꢀ 70 °C. Reaction was started by adding 50 µL of
Toruzyme 3.0L or 300 µL of αꢀglucosidase stock solution (134 U/mL). To determine the initial
reaction rates, Toruzyme 3.0L was diluted 162ꢀfold and 50 µL of diluted enzyme were added to
the reaction mixture. The total reaction volume was always 1 mL. The tubes were incubated in a
thermomixer at the specified temperature at an agitation rate of 300 rpm in the dark. The reaction
mixture was equilibrated for 10 min to the desired temperature before adding the enzyme.
Samples were withdrawn at certain times and the reaction was quenched by diluting with 1.2 M
HCl to a final acid concentration of 100 mM. After 10 min, the samples were neutralized with
dilute aqueous NaOH. The samples were appropriately diluted and subsequently analyzed by
HPLC for qualitative and quantitative determination of the
formed.
LꢀAA transglucosylation products
Trimming of Oligoglucosylated Forms of L-AA by Glucoamylase: Reaction mixtures (2
mL) from Toruzymeꢀcatalyzed transglucosylation for 24 h were divided in two parts. One part (1
mL) was heated to 95°C for 30 min to inactivate the CGTase. The other part was kept on ice
during that time and served as a control. Both samples were brought to 60 °C. The sample that
ACS Paragon Plus Environment
6

Page 7 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
had undergone heat treatment was supplied with 50 µL of glucoamylase stock solution (1000
U/mL). The untreated sample received the same volume of buffer. Both samples were incubated
for 2 h at 300 rpm and in the dark. They were then analyzed by HPLC as described above. Thus
the effect of oligoglucosyl chain trimming was determined by glucoamylase. The control was
designed to evaluate the intrinsic trimming activity of Toruzyme.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Isolation of AA-2G: A procedure adapted from Aga et al.19 was used. The glucoamylaseꢀ
treated reaction mixture (~140 g/L AAꢀ2G; 10 mL) was loaded onto an anionꢀexchange
chromatography column (XK 50/30, GE Healthcare Life Sciences, Vienna, Austria; bed volume
of 150 mL) comprising a Dowex 1X2 (chloride form) resin packed at a flow rate of 15 mL/min
and 0.3 MPa. Elution was performed at the same flow rate, starting with water (1.2 l) followed
by three steps of NaCl concentration (0.1 M, 0.25 l; 0.25 M, 0.15 l; 0.5 M, 0.20 l). Glucose and
other uncharged carbohydrates were eluted with water. AAꢀ2G and LꢀAA eluted together at 0.1
and 0.25 M NaCl. Fractions containing AAꢀ2G were pooled and concentrated 10ꢀfold in vacuo at
40 °C. The sample was then loaded onto a column (XK 50/30, bed volume of 150 mL) packed
with a 1:1 mixture of activated charcoal and calcined Celite 501. Elution was performed with a
flow rate of 15 mL/min employing a threeꢀstep gradient starting with water (0.5 l) followed by
10% (0.5 l) and 25% (1.0 l) of ethanol. Salts were eluted with 10% ethanol. Fractions eluted at
25% ethanol contained most of the AAꢀ2G. They were pooled and concentrated to oneꢀtenth of
volume in vacuo. After lyophilization, pure AA2G (98%, by HPLC) was obtained as a white
powder. Its structure was confirmed by 1H and 13C NMR analyses.
Transglucosylation Product Analysis: LꢀAA and glucosylated forms of LꢀAA were analyzed
using HPLC coupled to UV detection at 243 nm. A Shimadzu (Tokyo, Japan) HPLC 20A system
equipped with a UVꢀVis detector was used. Separation was performed with a Lichrosper 100
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 37
1
2
3
4
5
6
7
8
9
C18 reverseꢀphase column (250 × 4.6 mm, 5 µm; MachereyꢀNagel; Düren, Germany) operated at
30 °C. Gradient elution was performed at a constant flow rate of 1 mL/min using potassium
phosphate buffer (10 mM; pH 2.0) that contained varied amounts of acetonitrile (solvent B) in
different steps: 0ꢀ2 min B = 0%; 2ꢀ10 min B = 2%; 10ꢀ20 min B = 0%. Separation of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
oligoglucosylated
Authentic standards of
of ꢀAA were quantified based directly on the UV response factor of AAꢀ2G.
The transglucosylation products were also analyzed and their masses identified by LCꢀMS.
L
ꢀAA having side chains of up to 12 glucosyl residues could thus be achieved.
L
ꢀAA and AAꢀ2G were used for quantification. Oligoglucosylated forms
L
Sample analysis was conducted with an Agilent Technologies (Santa Clara, CA, USA) 1200
HPLC system coupled to a 6120 Quadrupole mass spectrometer (also from Agilent
Technologies) equipped with an electro spray ionization source. The separation was performed
using the Lichrosper column under the conditions described above, except that 10 mM
ammonium acetate supplemented with 0.5 mM LiCl (pH 6.75) replaced the phosphate buffer.
The mass spectrometer was used under the following instrument settings: athmospheric pressure
ionizationꢀelectro spray positive scan mode, 10.5 L/min drying gas flow (N2), 35 psig nebulizer
pressure, 350 °C drying gas temperature, 3 kV capillary voltage and 30 V fragmentor voltage.
Determination of Site-selectivity of Transglucosylation: Although LꢀAA offers a total of
four hydroxy groups as potential glucosylation sites, it was known from prior work in the
field19,20,29 that enzymatic transglucosylations give AAꢀ6G, AAꢀ2G or both. To distinguish
between AAꢀ2G and AAꢀ6G (Scheme 2), an HPLC analysis was used that employed an
Aminex® HPXꢀ87H ion exclusion column (300 × 7.8 mm; BioꢀRad, Hercules, CA, USA)
operated at 25 °C. The UV detection described above was used. An isocratic elution with 20 mM
H2SO4 in water was used. A constant flow rate of 0.4 mL/min was applied. Samples of
ACS Paragon Plus Environment
8

Page 9 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
Toruzymeꢀcatalyzed reactions underwent treatment with glucoamylase to eliminate effects of
longer oligoglucosyl side chains on product separation and identification. Samples from
reactions of the αꢀglucosidase from rice (known to produce both AAꢀ2G and AAꢀ6G)20,19 and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
the αꢀglucosidase from A. niger (known to produce mainly AAꢀ6G)40 were used without
treatment.
NMR spectroscopy: A Varian (Agilent Technologies) INOVA 500ꢀMHz NMR spectrometer
1
and the VnmrJ 2.2D software were used for all measurements. H NMR spectra (499.98 MHz)
were measured on a 5 min indirect detection pulsedꢀfieldꢀgradient probe, while a 5 mm dualꢀ
directꢀdetection probe with zꢀgradients was used for collecting 13C NMR spectra (125.71 MHz).
RESULTS AND DISCUSSION
High-temperature Reaction Conditions For AA-2-G Synthesis by Toruzyme: Despite
previous studies of Toruzyme41,42 it was necessary to find suitable reaction conditions for AAꢀ2G
synthesis.
LꢀAA stability was considered important in particular. Reported syntheses of AAꢀ2G
using CGTase29 or another transglucosidase14 differed in whether thiourea was added to prevent
oxidative degradation of
activity of Toruzyme and the stability of
Figure 2 (panel A) shows the effect of temperature in the range 30 to 70 °C on the amounts of
AAꢀ2G formed and ꢀAA consumed in reactions with and without thiourea (2.6 mM). At
LꢀAA. An acidic pH (4.5) was chosen because the transglycosylase
LꢀAA were both high under these conditions.
L
temperatures of up to 50 °C, the AAꢀ2G formation was enhanced by about 20 to 30% on addition
of thiourea. The benefit of thiourea vanished gradually at the higher temperatures. When thiourea
was absent, the LꢀAA consumption exceeded the AAꢀ2G release at all temperatures, suggesting
loss of
L
ꢀAA due to instability. On addition of thiourea, a close mass balance for the two
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 37
1
2
3
4
5
6
7
8
9
compounds was obtained in the temperature range 30 ꢀ 50 °C. At 60 °C and higher, however,
effect of thiourea on was strongly attenuated. Because the highest AAꢀ2G amounts were
obtained at 50 °C (Figure 2, panel A), all further experiments were performed at this temperature
in the presence of thiourea.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Reaction Time Course Analysis: A typical time course of AAꢀ2G synthesis is displayed in
Figure 2 (panel B). Initially in the reaction, there was concurrent release of AAꢀ2G and different
oligoglucosylated forms of LꢀAA, in short AAꢀ2Gn+1 where n ≥ 1 is the number of additional
glucosyl residues attached. We discuss later the implication that n was observed to be as high as
11, despite the fact that a donor substrate comprising only 6 glucosyl residues was applied.
Figure 2 (panel B) shows that the summedꢀup amount of AAꢀ2Gn+1 passed through a maximum
shortly after the start of the reaction, only to decrease slowly afterwards. By contrast, AAꢀ2G
concentration increased in a more gradual manner, levelling out at a time when the AAꢀ2Gn+1
products had already been depleted from the reaction mixture to a substantial degree. HPLC
analysis that resolved
LꢀAA oligoglucosides according to their nꢀvalue (Figure 3) revealed
Toruzymeꢀcatalyzed degradation of the AAꢀ2Gn+1 products through "exoꢀmode" hydrolysis of
their oligoglucosyl chains. (Note: MS analysis was used to confirm the masses of AAꢀ2Gn+1
products (n ≤ 5), as shown in Figure S1 of Supporting Information.) Longer chain products
disappeared faster from the reaction mixture than others comprising smaller chains, as shown in
Figure 3. Even though conversion of AAꢀ2Gn+1 into AAꢀ2G by Toruzymeꢀintrinsic hydrolase
activity was relatively slow, it was suggested from literature19 and confirmed in our experiments
(see later) that postꢀsynthetic processing of AAꢀ2Gn+1 by addition of glucoamylase was highly
effective. Therefore, optimization for conditions of synthesis focused on maximizing not only the
ACS Paragon Plus Environment
10

Page 11 of 37
ACS Catalysis
1
2
3
4
amount of AAꢀ2G6 but also the total amount of transglucosylation product, that is, AAꢀ2G plus
5
6
AAꢀ2Gn+1
.
7
8
9
Effect of Acceptor and Donor Concentration, and of the Acceptor/Donor Ratio: In
enzymatic transglucosylations, the acceptor is commonly used in a suitable molar excess over the
donor to suppress the competing hydrolysis.43 To define a useful "operational window" for AAꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2G synthesis, experiments were performed at a variable molar ratio
range 1.4 to 22. The total initial concentration of αꢀcyclodextrin and
g/L. Product formation after 24 h exhibited a broad plateau at ꢀAA/αꢀcylcodextrin ratios
L
ꢀAA/αꢀcyclodextrin in the
LꢀAA was constant at 37
L
between around 4 and 13 (Figure S2, Supporting Information). Analysis of the initial rates of
transglucosylation gave a similar result, as shown in Figure S3 and Table S1 in Supporting
Information. Because utilization of the acceptor becomes progressively less efficient as the
acceptor/donor ratio rises, we chose a ratio at the lower end (
range explored for the following experiments.
LꢀAA/αꢀcylcodextrin = 4.6) of the
Figure 4 shows the effect of varying αꢀcyclodextrin concentration in the range 31 – 310 mM
on the formation of AAꢀ2G (panel A) and total AAꢀ2Gn+1 products (panel B) over time.
Production was enhanced through increase in donor substrate concentration with the highest AAꢀ
2G output being obtained when initial αꢀcyclodextrin concentration was 248 mM. A greater
initial αꢀcyclodextrin concentration brought about no further increase. However, the highest
amount of total AAꢀ2Gn+1 products was obtained at an initial αꢀcyclodextrin concentration of
310 mM. In contrast to AAꢀ2G formation, which except for reaction at low αꢀcyclodextrin
concentration increased steadily over time (Figure 4, panel A), total AAꢀ2Gn+1 concentration
increased only shortly into the reaction (≤ 6 h) and stopped sharply afterwards (Figure 4, panel
B). On the other hand, AAꢀ2G formation continued, albeit at a rate that was significantly slower
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 37
1
2
3
4
5
6
7
8
9
than in the early reaction phase. A plausible explanation for this behavior is that: (a) AAꢀ2G and
other AAꢀ2Gn+1 products were synthesized only early in the reaction; (b) any further increase in
AAꢀ2G results from degradation of the oligoglucosylated chains in AAꢀ2Gn+1 by the intrinsic
hydrolase activity of CGTase. The molar product ratio ΣAAꢀ2Gn+1/AAꢀ2G after exhaustive
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
reaction (24 h) increased from a value of 0.15 at a low αꢀcyclodextrin concentration to a value of
1.03 at a high αꢀcyclodextrin concentration, suggesting that elevated donor substrate
concentrations drive the overall synthesis towards oligoglucosylation of LꢀAA.
Trimming of the Oligoglucosyl Chain in AA-2Gn+1 Products by Glucoamylase Boosts the
AA-2G Yield: It was known from literature,19 and was also suggested by results of this study
(see Figure 2, panel B) that AAꢀ2Gn+1 products can be converted to AAꢀ2G through hydrolysis
of their oligoglucosyl chains. Rather than relying on the low hydrolase activity of the CGTase
itself, use of a second hydrolase has the advantage of speeding up the process considerably. Here
the glucoamylase from A. niger was used. The enzyme is a known exoꢀhydrolase that degrades
αꢀ1,4ꢀoligoglucosyl chains in a processive manner, cleaving glucose residues from the nonꢀ
reducing chain end.19,39 AAꢀ2G was shown here not to be a substrate of the glucoamylase. Figure
5 (panel A) shows HPLC traces that illustrate the composition of a product mixture before and
after glucoamylase treatment. It is shown that all peaks of AAꢀ2Gn+1 products disappeared and
the peak of AAꢀ2G increased strongly on incubation with the glucoamylase. Panel B of the same
Figure 5 shows the impact of the glucoamylase "oligoglucosyl chain trimming" on AAꢀ2G
recovery from different transglucosylation reactions. As expected from the relative abundance of
oligoglucosylated LꢀAA present in the mixtures (see Figure 4), the glucoamylase trimming was
relatively more effective in reactions where a high αꢀcyclodextrin concentration had been used.
It was possible to double the AAꢀ2G concentration by means of the glucoamylase treatment
ACS Paragon Plus Environment
12

Page 13 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
(Figure 5, panel B). Notably, the amount of AAꢀ2G thus obtained exceeded on a molar basis that
of the αꢀcyclodextrin employed. For example, a reaction starting from 310 mM αꢀcyclodextrin
furnished 424 mM AAꢀ2G, demonstrating the effect to be substantial and also highly significant
regarding the overall AAꢀ2G yield from αꢀcyclodextrin. This reflects boost of the donor
substrate utilization by a factor of 1.37 (= 424/310) compared to transglucosylation by coupling
only. Note: the theoretical molar yield of coupling is 1. Therefore, a CGTase having high
disproportionation activity should be a very useful catalyst for a yieldꢀenhanced synthesis of AAꢀ
2G. Using αꢀcyclodextrin concentrations lower than 310 mM this enhancement decreased
gradually but was significant in all reactions, as shown in Figure 5.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Site-selectivity of Glucosylation of
L-AA by Toruzyme: Figure 6 shows results of an
analysis to characterize the site selectivity of Toruzymeꢀcatalyzed glucosylation of
LꢀAA under
the conditions used. Contrary to reactions catalyzed by αꢀglucosidases that gave the wrong
product regioisomer AAꢀ6G (A. niger enzyme) or a mixture of AAꢀ2G and AAꢀ6G (rice
enzyme) as expected, absence of AAꢀ6G from reactions of Toruzyme was demonstrated within
detection limits. Therefore, Toruzyme presents a highly siteꢀselective catalyst for the formation
of 2ꢀOꢀαꢀ
in Figure 5 (panel A), AAꢀ2G was obtained as a single glucosyl transfer product from
Characterization of Glucosylation of -AA via Disproportionation: The distribution of
oligoglucosylated ꢀAA products in samples from the transglucosylation at extended reaction
D
ꢀglucosyl linkages on glucosylation of
LꢀAA. After glucoamylase treatment, as shown
L
ꢀAA.
L
L
time reflects the complex interplay between the different reactions catalyzed by the CGTase
(Scheme 1). To distinguish between product formation due to direct coupling and
disproportionation, it was therefore necessary to analyze the reaction at an early stage (≤ 30 min)
where the distribution of transglucosylation products showed strong variation with time and was
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 37
1
2
3
4
5
6
7
8
9
not primarily shaped by secondary conversion processes. Figure 7 (panel A) reveals that besides
AAꢀ2G, which was the predominant glucosylation product at all times, oligoglucosylated forms
(AAꢀ2G2 ꢀ AAꢀ2G12) were also present in appreciable amounts after just 10 min. Reactions
performed under initial rate conditions (Figure S3 and Table S1, Supporting Information)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
confirmed the presence of different oligoglucosylated forms of
the conversion. The largest transglucosylation product was AAꢀ2G6 under these conditions and it
accounted for about oneꢀsixth to oneꢀfifth of the total amount of ꢀAA glucosides formed (Table
LꢀAA right from the beginning of
L
S1, Supporting Information). However, already very early in the reaction (Figure S3, Supporting
Information) the release of AAꢀ2G6 leveled off and started to decrease even. Therefore, these
results clearly indicated that direct coupling, to produce AAꢀ2G6 (Scheme 1), was not the
preferred reaction path of glucosylation of
LꢀAA. Instead it seemed that disproportionation,
probably involving maltohexaose formed through rapid hydrolysis of αꢀcyclodextrin, constituted
the main reaction path, as shown in Scheme 3.
Evidence in Figure 7 (panel B) confirmed maltohexaose to be active as donor substrate of
Toruzymeꢀcatalyzed glucosylation of
LꢀAA. However, the chainꢀlength distribution of
transglucosylation products from maltohexaose was much steeper, that is, shifted strongly
toward products having smaller chains, as compared to reaction using αꢀcyclodextrin. Product
distribution obtained from maltohexaose was independent of reaction time whereas in the case of
αꢀcyclodextrin, the distribution appeared somewhat bimodal in the beginning (10 ꢀ 20 min) with
maxima at AAꢀ2G and AAꢀ2G6, only to shift later to a unimodal distribution with a maximum at
AAꢀ2G. However, even at a longer time (1 h), a decline in the relative abundance from AAꢀ2G
to longerꢀchain transglucosylation products remained less pronounced in the reaction with αꢀ
cyclodextrin as compared to reaction with maltohexaose. We also noted that reaction from
ACS Paragon Plus Environment
14

Page 15 of 37
ACS Catalysis
1
2
3
4
maltohexaose was somewhat (about 6.5ꢀfold) slower than that from αꢀcyclodextrin and produced
5
a smaller (about 3.6ꢀfold) amount of transglucosylation products. Therefore, this suggested that a
6
7
8
9
"secondary disproportionation" where oligoglucosylated
LꢀAA products served as donor
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
substrates for another glucosylation of ꢀAA also contributed to product formation and to the
L
shaping of the product distribution obtained from αꢀcyclodextrin (Figure 7, panel A). Secondary
disproportionation provides a reasonable explanation for the high preponderance of AAꢀ2G as
transglucosylation product, as follows. CGTase is thought to cleave linear donor substrates with
release of G2 units from the reducing end.44,45 Assuming the same cleavage pattern for
glucosides whereby ꢀAA is recognized as the reducingꢀend residue, AAꢀ2G6 for example would
be converted into AAꢀ2G and the remaining maltopentosyl moiety might be transferred to
another ꢀAA acceptor to give AAꢀ2G5. Knowing the product distribution due to
disproportionating transglucosylation from maltohexaose (Figure 7, panel B), it was possible to
roughly estimate that the relative contributions of direct coupling to formation of ꢀAA
glucosides from αꢀcyclodextrin was only about oneꢀsixth that of disproportionation (Scheme 3).
Note that direct coupling to ꢀAA gives AAꢀ2G6 as a single product. The remaining products are
LꢀAA
L
L
L
L
formed through disproportionation reactions.
As already shown in Figure 4 (panel B), the formation of LꢀAA transglucosylation products
except that of AAꢀ2G proceeded only for a few hours until oligoglucosyl chain hydrolysis
became the predominant reaction. Hydrolysis was estimated to have been 20 times slower than
the transglucosylation. More detailed analysis of the product distribution in the "hydrolysis
phase" (6 ꢀ 24 h; data not shown) revealed that, shorterꢀchain oligoglucosides (n ≤ 3) were
observed to accumulate apparently at the expense of products having longer side chains.
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 37
1
2
3
4
Toruzyme Prefers AA-2G Over
L-AA As Acceptor Substrate For Transglucosylation: It
5
6
7
8
was known from earlier studies38,45–47 that Toruzyme utilizes different glucosides (e.g.
steviosides, steroidal saponins and others) as acceptor substrates, thus converting them into
oligoglucosylated products. To determine if glucosylation of AAꢀ2G contributed to the amount
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
and the distribution of
experiment in which both
L
ꢀAA glucosides formed, we performed a "competing acceptor"
L
ꢀAA (142 mM) and AAꢀ2G (24 mM) were present at start of the
reaction to become glucosylated from αꢀcyclodextrin. A control experiment lacking AAꢀ2G was
performed. Figure 8 shows the ꢀAA glucoside product distribution in samples from the two
L
reactions. When AAꢀ2G was initially present, the abundance of oligoglucoside products having n
≥ 6 was enhanced in a pronounced manner compared to the control. The relative amount of
products having 6 to 8 glucosyl residues increased up to eightfold (Figure 8, panel B). Timeꢀ
resolved analysis (data not shown) revealed that the AAꢀ2G concentration dropped within the
first 6 h from the initial value of 24 mM to 15 mM, only to remain constant afterwards until 24 h.
About 11 mM AAꢀ2G was produced in the control. These results implicate AAꢀ2G as acceptor
substrate for Toruzymeꢀcatalyzed transglucosylation and they also reveal clear preference of the
enzyme for glucosylation of AAꢀ2G compared to glucosylation of LꢀAA. They also show the
potential for tuning the composition of the product mixture towards longerꢀchain oligoglucosides
(Figure 8, panel B). It appears that AAꢀ2G competes much more efficiently with water for
reaction with the glucosylated CGTase than LꢀAA is able to, thus promoting the formation of
longꢀchain oligoglucosylated products through enhanced direct coupling from αꢀcyclodextrin
and also "primary" disproportionation reactions. This is the first report describing
transglucosylation of AAꢀ2G with detailed product analysis.
ACS Paragon Plus Environment
16

Page 17 of 37
ACS Catalysis
1
2
3
4
The Benefit of Disproportionation - Highly Efficient Synthesis of AA-2G by Toruzyme-
5
6
7
Catalyzed Transglucosylation From α-Cyclodextrin: A summary of the reactions involved in
8
9
the Toruzymeꢀcatalyzed transglucosylation of LꢀAA is shown in Scheme 3. Through effectively
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
combining coupling with disproportionation, Toruzyme enabled a highly efficient utilization of
αꢀcyclodextrin as donor substrate for AAꢀ2G production. Table 1 enables a comparison of the
results from this and earlier studies of AAꢀ2G synthesis by transglucosylation. CGTases from
Bacillus sp. were recognized for their relatively high coupling activity from αꢀcyclodextrin, and
the enzyme from B. stearothermophilus was regarded as the most suitable biocatalyst for
commercial AAꢀ2G production. To replace cyclodextrin with a less expensive and in terms of
glucosyl content more completely utilizable donor substrate (e.g. starch, maltodextrin, maltose),
other CGTases in native or engineered form were considered. Variants of Paenibacillus
macerans CGTase showed enhanced specificity towards these nonꢀcyclic substrates.30ꢀ32
However, the AAꢀ2G yields thus obtained were still much lower than the "benchmark" yield
using cyclodextrin.30,32,48,49 Interestingly, the increased disproportionation activity of the CGTase
variants proved beneficial for the AAꢀ2G yield from cyclodextrin.32,50,51 However, the
engineered CGTases often showed a substantially lowered efficiency in the reaction with
LꢀAA
as compared to the wildꢀtype enzyme. Table 1 indicates strong variation in the degree of
utilization of
utilization of
L
ꢀAA in the different studies on AAꢀ2G synthesis performed. With a 30%
ꢀAA, the enzymatic reactions performed herein appear to be relatively efficient.
ꢀAA offered. Recent developments
L
CGTase variants utilized just a few percent of the
L
notwithstanding, there appears to be still a long way to an optimally engineered CGTase for
efficient AAꢀ2G synthesis.
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 37
1
2
3
4
5
6
7
8
9
Advances made with these alternative systems notwithstanding, transglucosylation from αꢀ
cyclodextrin remains the strongest candidate for CGTaseꢀcatalyzed synthesis of AAꢀ2G (Table
1). An important limitation of the previous studies was their exclusive focus on optimization of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
the AAꢀ2G concentration, thus not considering the formation of
importance of these "AAꢀ2G precursor products" for an optimized utilization of the donor
substrate. The current study is the first to take into account the total amount of ꢀAA glucosides
LꢀAA oligoglucosides and the
L
formed in the reaction as a target for process optimization and to recognize the significant
potential of CGTaseꢀcatalyzed primary and secondary disproportionations for enhanced AAꢀ2G
production. Thus, to our knowledge, this study discloses AAꢀ2G concentrations that exceed
those so far reported, obtaining them by means of practical conditions and consideration of the
mass ratio of donor and
enzymes and the above process conditions can be used for largeꢀscale production of AAꢀ2G.
After glucoamylase treatment the reaction mass contains AAꢀ2G along with unreacted
LꢀAA used. Toruzyme and glucoamylase are commercially available
LꢀAA
and sugars, mostly glucose. Using anion exchange and activated carbon chromatographies, we
isolated AAꢀ2G at gram scale in ≥98% purity with a yield of 52% or higher. The expected
product structure was confirmed by 1H and 13C NMR (Figure S4, Supporting Information).
CONCLUSIONS
Exploiting the high disproportionation activity of Toruzyme, the efficiency of CGTaseꢀcatalyzed
synthesis of AAꢀ2G by transglucosylation from αꢀcyclodextrin could be enhanced significantly
in terms of both product concentration and degree of utilization of donor substrate. The improved
use of αꢀcyclodextrin is considered especially in reducing the cost contribution of the donor
substrate to total production cost. Using detailed analysis of the distribution of LꢀAA glucosides
ACS Paragon Plus Environment
18

Page 19 of 37
ACS Catalysis
1
2
3
4
5
6
formed in the reaction over time, reaction paths involved in the transglucosylation were revealed
and their relative importance to the formation of product was identified.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 37
1
2
3
4
FIGURES
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. The chemical structures of
L
ꢀascorbic acid (
L
ꢀAA) (A), 2ꢀOꢀαꢀDꢀglucopyranosylꢀLꢀ
ascorbic acid (AAꢀ2G) (B), and αꢀcyclodextrin (C) are shown.
ACS Paragon Plus Environment
20

Page 21 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Toruzymeꢀcatalyzed glucosylation of
L
ꢀAA is shown. A. Temperature dependence of
AAꢀ2G formation (circles) and the corresponding
L
ꢀAA depletion (triangles) in the absence
(closed symbols) and presence (open symbols) of 2.6 mM thiourea. The reaction time was 4 h. B.
Time course of AAꢀ2G (closed circles) and AAꢀ2Gn+1 (open circles) production during reaction
at 50 °C in the presence of 2.6 mM thiourea. All reactions were performed at pH 4.5 and used 30
mM αꢀcyclodextrin, 135 mM LꢀAA, and a Toruzyme loading of 50 µL/mL. The symbols show
the measured data. The lines indicate the data trend. Results are mean values of at least triplicate
determination and have an S.D. of 6% or smaller.
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 37
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Degradation of transglucosylation products (AAꢀ2Gn+1 where n ≥ 1) through a low
"exoꢀtrimming" activity of Toruzyme is shown. The HPLC traces show the product composition
after 3 h and 24 h of reaction using conditions of Figure 1 (panel B). Peaks of AAꢀ2Gn+1
products were designated with numbers according to their nꢀvalues.
ACS Paragon Plus Environment
22

Page 23 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Time courses of AAꢀ2G (panel A) and total AAꢀ2Gn+1 (panel B) formation at the
different αꢀcyclodextrin concentrations (mM) indicated using a constant ꢀAA/ αꢀcyclodextrin
L
ratio of 4.6. The AAꢀ2Gn+1 products analyzed and included in panel B have nꢀvalue between 1
and 11. Reactions conditions: 50 µL Toruzyme/mL, 2.6 mM thiourea, pH 4.5 and 50 °C. The
symbols show the measured data and the dotted lines indicate the data trend. Results are mean
values of at least triplicate determination and have an S.D. of 6% or smaller.
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 37
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Effect of glucoamylase (GA) treatment on the recovery of AAꢀ2G from AAꢀ2Gn+1
transglucosylation products is shown. A. Superimposition of HPLC traces of reaction sample
before (control) and after the GA treatment. Peak numbering 1 – 3 indicates the value of n in
AAꢀ2Gn+1. B. Change in AAꢀ2G concentrations in GAꢀtreated and untreated (control) samples
obtained from transglucosylations performed at different αꢀcyclodextrin concentrations (mM) as
indicated. AAꢀ2Gn+1 formation was done for 24 h under the reaction conditions of Figure 3. One
sample was treated with GA (50 U/mL) for 2 h. The control was incubated identically without
GA.
ACS Paragon Plus Environment
24

Page 25 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Superimposition of HPLC traces from the analysis of reaction samples of
transglucosylations catalyzed by Toruzyme and the αꢀglucosidases (αꢀGase) from rice and A.
niger. The trace of an authentic standard of AAꢀ2G is also shown. The enzyme from A. niger is
known to synthesize mainly AAꢀ6G whereas the rice enzyme produces both AAꢀ2G and AAꢀ6G.
Thus absence of AAꢀ6G from Toruzymeꢀcatalyzed reactions was shown clearly.
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 37
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Distribution of
L
ꢀAA glucosides formed in transglucosylations from αꢀcyclodextrin
(panel A) and maltohexaose (panel B) at early reaction times. The reaction conditions of Figure 3
were used. Reactions used 310 mM donor substrate and 1420 mM LꢀAA.
ACS Paragon Plus Environment
26

Page 27 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. The acceptor specificity of Toruzyme in the transglucosylation reaction is shown. A.
HPLC traces showing the effect of externally added AAꢀ2G at reaction start on the distribution
of transglucosylation products formed within 6 h. B. Relative increase in the concentration of
each transglucosylation product due to the initial presence of AAꢀ2G. Reactions were performed
under conditions of Figure 1 (panel B) using 31 mM αꢀcyclodextrin and 142 mM LꢀAA in the
absence (control) and presence of AAꢀ2G (24 mM).
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 37
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Tranglucosylation catalyzed by CGTases. A. Double displacementꢀlike mechanism of
the reaction (transglucosylation to a suitable acceptor, hydrolysis). B. Reaction pathways via
cyclization, coupling, disproportionation and hydrolysis. Blue circles show glucosyl residues, the
white circle is the reducingꢀend glucose or free glucose. The acceptor substrate glucosylated in
the reaction is shown as a yellow frame. The red line indicates an enzymatic cleavage site.
Maltose is a preferred cleavage product in the first (glucosylation) step of the reaction of
CGTase, as shown in panel A.
ACS Paragon Plus Environment
28

Page 29 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Synthesis of AAꢀ2G by biocatalytic transglucosylation to
LꢀAA (1) requires a
suitably siteꢀselective enzyme. Transglucosylation from the disaccharide maltose gives a product
mixture of AAꢀ6G (2) and AAꢀ2G (3) when the rice αꢀglucosidase (αꢀGase) is used or mainly
AAꢀ6G when the A. niger αꢀglucosidase (αꢀGase) is used. Use of CGTase (Toruzyme) in
combination with αꢀcyclodextrin as donor substrate could result in a completely siteꢀselective
glucosylation of
glucopyranosylꢀ
L
ꢀAA. 2 – 6ꢀOꢀαꢀ
D
ꢀglucopyranosylꢀ
Lꢀascorbic acid (AAꢀ6G); 3 – 2ꢀOꢀαꢀDꢀ
L
ꢀascorbic acid (AAꢀ2G)
ACS Paragon Plus Environment
29