A high-yielding biocatalytic process for the production of 2-O-(α-D-glucopyranosyl)-sn-glycerol, a natural osmolyte and useful moisturizing ingredient

Article abstract of DOI:10.1002/anie.200803562

(Figure Presented) Fit for industry: Stereochemically pure 2-O-(α-D-glucopyranosyl)-sn-glycerol (αGG) was obtained in high yield from an efficient and selective biocatalytic process (see schematic outline). The sucrose phosphorylase catalyzed transfer of a glucosyl group from sucrose to glycerol unites the main advantages of transglycosidases, glycosyltransferases, and glycosynthases for glycoside synthesis and provides access to αGG as an industrial chemical.

Full text of DOI:10.1002/anie.200803562

Communications
DOI: 10.1002/anie.200803562
Applied Biocatalysis
A High-Yielding Biocatalytic Process for the Production of
2-O-(a-d-glucopyranosyl)-sn-glycerol, a Natural Osmolyte and Useful
Moisturizing Ingredient**
Christiane Goedl, Thornthan Sawangwan, Mario Mueller, Alexandra Schwarz, and
Bernd Nidetzky*
Dedicated to Professor Herfried Griengl on the occasion of his 70th birthday
Compatible solutes constitute a diverse class of small organic
molecules, many of which have a glycosidic chemical struc-
ture. They are found ubiquitously in nature, where they serve
an essential function in protecting cells against high salt
concentration, extremes of temperature, and other forms of
external stress. Their physiological effectiveness[1] and tech-
nological performance[2] can be traced back to their ability to
regulate the cellular water balance, prevent protein denatu-
ration, and stabilize supramolecular biological structures,
such as those of lipid membranes.
productivity, or a combination of these problems.[6,8] We
describe herein a new biocatalytic process which overcomes
the chemical and technological challenges associated with the
production of stereochemically pure aGG as an industrial
chemical.
Scheme 1 summarizes the key features of the enzymatic
synthesis of aGG by transglucosylation to glycerol from
sucrose. Under natural conditions, sucrose phosphorylase
(EC 2.4.1.7) catalyzes the reversible conversion of sucrose
and phosphate into a-d-glucose 1-phosphate (aG1P) and
Glycosyl glycerols are powerful osmolytes that are
produced by various plants, algae, and bacteria in adaptation
to salt stress and drought.[3] Among them, 2-O-(a-d-gluco-
pyranosyl)-sn-glycerol (aGG), the main compatible solute in
Scheme 1. Comparison of enzymatic routes for the production of aGG
by a) transglucosylation and b) a hypothetical “a-glucosynthase”-cata-
lyzed reaction. The relative flux through each step is indicated by arrow
length and derived from kinetic data from this study and reference [6].
Available kinetic data for a b-glucosynthase-catalyzed reaction were
used for path b.[25] The problems of regioselectivity (dotted lines) and
secondary product hydrolysis (c) are indicated. 1-aGG, 1-O-(a-d-
glucopyranosyl)-sn-glycerol.
photosynthetic bacteria,[4] has attracted special attention as a
promising moisturizing agent in cosmetics,[5] but also as a low-
calorie sweetener for the prevention of tooth decay.[6] Possible
therapeutic applications based on the ability of aGG to
stabilize proteins and cells are currently under evaluation.[7]
However, the development of industrial applications for
aGG is severely restricted by compound availability.
Reported synthetic procedures are not technologically
mature as a result of insufficient yield, selectivity, or
d-fructose.[9] In the absence of phosphate, glycerol can
intercept the b-glucosyl enzyme intermediate of the reaction
with sucrose to produce aGG; hydrolysis of the glucosylated
enzyme can also occur as a side reaction.[10] Although the
overall strategy of the transglycosidase-catalyzed synthesis of
glycosides is well established in carbohydrate chemistry,[11]
several peculiarities make this biocatalytic process to our
knowledge unique. First of all, the active site of sucrose
phosphorylase provides splendid control over the regioselec-
tivity of glucosyl transfer. The regioselectivity of this process
is often insufficient with other transglycosidases (see refer-
ence [12] for the general case and reference [13] for the
[*] C. Goedl, T. Sawangwan, M. Mueller, Dr. A. Schwarz,
Prof. Dr. B. Nidetzky
Institute of Biotechnology and Biochemical Engineering
Graz University of Technology
Petersgasse 12/1, 8010 Graz (Austria)
Fax: (+43)316-873-8434
E-mail: bernd.nidetzky@tugraz.at
Homepage: http://www.biote.tugraz.at
[**] Financial support from the FWF Austrian Science Fund (DK
Molecular Enzymology W901-B05) and the OEAD is gratefully
acknowledged. We thank Prof. Dr. Lothar Brecker, University of
Vienna, for the NMR spectroscopic analysis.
10086
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 10086 –10089

Angewandte
Chemie
synthesis of aGG). Second, with the enzyme sucrose phos-
phorylase, the competing reaction with water is suppressed
kinetically to the extent that hydrolysis of the substrate is
prevented completely in the presence of glycerol as an
acceptor at a suitable concentration. Third, the use of sucrose
as a high-energy glucosyl donor, in combination with the
substantial kinetic hindrance to the degradation of aGG by
sucrose phosphorylase, provides a large driving force for an
essentially unidirectional reaction, which gives the product in
almost quantitative yield (on the basis of converted sub-
strate). Therefore, the biocatalytic process for the production
of aGG unites the main synthetic advantages of transglyco-
sidases (with respect to the relative simplicity of the reaction
system and the use of cheap substrates), glycosyltransferases
(with respect to product uniformity due to the high regiose-
lectivity of glucosyl transfer), and glycosynthases (with
respect to the kinetic stability of the glycosidic product).[14,15]
Steady-state kinetic assays were performed to study the
competing reactions of the glucosyl enzyme intermediate of
sucrose phosphorylase with glycerol and bulk water. Depend-
ing on the glucosyl donor used as the substrate, the rate of
fructose (VFru ; sucrose) or phosphate (VPi; aG1P) release was
measured along with the rate of glucose formation (VGlc). The
ratios VFru/VGlc and VPi/VGlc were determined at varying
concentrations of glycerol (Figure 1a). With each donor
substrate, the rate ratio increased with an increasing concen-
tration of glycerol, as expected if glycerol competes with bulk
water to react with the glucosylated enzyme (Scheme 1). Fits
of straight lines to data obtained with sucrose and aG1P gave
slope values of 7.9mÀ1and 2.8mÀ1, respectively. These values
reflect the kinetic partition coefficient of the glucosylated
enzyme under the conditions used and indicate clearly that, in
contrast to previous findings,[16] glycerol takes part in the
enzymatic reaction as an acceptor of the glucosyl residue
transferred from the enzyme. Importantly, the leaving group
of the donor substrate influenced the overall efficiency of
glucosyl transfer to glycerol, a result which is not accounted
for by Scheme 1. We ascribe the observed effect tentatively to
conformational flexibility at the acceptor-binding site of
sucrose phosphorylase. Such flexibility is apparent from
high-resolution X-ray structures of the enzyme (from Bifido-
bacterium adolescentis).[17]
The ramifications of the kinetic evidence for the synthesis
of aGG are shown in Figure 1b, in which the course of
product formation over time is compared for the enzymatic
conversion of sucrose and aG1P. The yield of transfer product
as determined by HPLC was much higher with sucrose
(ca. 95% with respect to conversion of the donor substrate)
than with aG1P. The formation of free glucose serves as a
measure of the fraction of glucosyl residues not transferred to
glycerol. The data in Figure 1b corroborate the hypothesis
that hydrolysis competes more strongly with glucosyl transfer
when sucrose is replaced with aG1P as the donor substrate.
NMR spectroscopic analysis of the product mixture obtained
after the complete conversion of sucrose revealed that,
within the limits of detection of the methods used (0.02%),
only the desired regioisomer of a-glucosylglycerol had been
formed.
Table 1 summarizes the results of experiments carried out
to optimize the product and space–time yields of the
enzymatic conversion of sucrose and glycerol. Considering
the broad optimum pH range (5.0–8.0) for the enzymatic
formation of aG1P from sucrose,[18] there was an unexpect-
edly narrow operating range with respect to the pH value for
the synthesis of aGG. The optimum pH value is 7.0. By
varying the initial concentrations of sucrose and glycerol
systematically, we established conditions (0.8m sucrose, 2.0m
glycerol) under which the specific space–time yield (STY,
based on the amount of enzyme activity used per unit volume)
was highest and, at the same time, the product yield was
around 90%. A further increase in the concentrations of
sucrose and glycerol did not lead to significant improvements
in either the STY or the product yield. To determine the
required purity of the soluble biocatalyst, we compared the
performance of the crude Escherichia coli cell extract from
which recombinant sucrose phosphorylase was obtained with
that of the isolated enzyme. There was no detectable differ-
ence between the two enzyme preparations (Table 1). How-
ever, the use of whole bacterial cells that express sucrose
phosphorylase led to a substantial decrease in product yield as
compared to the same reaction with the free enzyme.
Figure 1. Comparison of sucrose and aG1P as glucosyl-donor sub-
strates for the synthesis of aGG. a) Kinetic partitioning analysis with
Figure 2 shows the complete course of aGG synthesis
over time under optimized reaction conditions. Importantly,
the hydrolysis of sucrose occurred to only a very small extent
throughout the reaction. The transfer product (0.7m) was
isolated with an estimated purity of ꢀ 98% (HPLC) and in an
overall yield of about 63% by a single-step chromatographic
*
*
sucrose (0.8m, ) and aG1P (0.1m, ) in the presence of sucÀro1 se
phosphorylase at concentrations of 20 UmLÀ1 ( ) and 3 UmL ( ).
*
*
*
b) Formation of aGG during the reaction of sucrose (0.3m, ) or
*
aG1P (0.1m, ) with glycerol (2.0m) in the presence of sucrose
phosphorylase (20 UmLÀ1) at 308C and pH 7.0.
Angew. Chem. Int. Ed. 2008, 47, 10086 –10089
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 10087

Communications
Table 1: Optimization of the synthesis of aGG with respect to the
Secondary product hydrolysis frequently presents a major
restriction to transglycosidase-catalyzed glycoside synthesis.
For this reason, the enzymatic reaction must usually be
performed in a tightly controlled kinetic regime, and product
yields are often only around 10–20%.[19] Concerns about
hydrolytic breakdown of the product have driven the devel-
opment of the glycosynthase concept. Glycosynthases are
engineered glycosidases that cannot promote secondary
hydrolysis because of structural changes in their active
site.[20] However, the successful uncoupling of product syn-
thesis and hydrolysis hinges on the use of highly activated
glycosyl donors, which are often chemically unstable.[14] In
light of the state of transglycosidase-catalyzed glycoside
synthesis, it was intriguing that the formation of aGG
appeared to be equilibrium controlled (see Figures 1b and
2), which implies an unusual kinetic stability of the product
under the transglucosylation conditions used. We therefore
measured the activity of sucrose phosphorylase by using
purified aGG as the glucosyl-donor substrate and phosphate,
water, or fructose as the acceptor. The turnover frequency
was extremely low in these reactions (kcat ꢁ 5 ꢀ 10À4 sÀ1) and
can be compared with the kcat values of about 100 sÀ1 and 2 sÀ1
for the conversion of sucrose into aG1P and d-glucose,
respectively. Thus, the replacement of d-fructose with glyc-
erol as the leaving group increased the stability of the
a-glucoside towards hydrolysis catalyzed by sucrose phos-
phorylase by nearly four orders of magnitude.
product yield and specific space–time yield (STY).[a]
c(sucrose) [m] c(glycerol) [m] Transfer yield [%] STY [mhÀ1 1000 kU]
0.3
0.5
0.8
1.0
1.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
93
93
85
80
55
47
0.54
0.97
1.41
1.11
1.15
1.31
0.8
0.8
0.8
0.8
0.8
0.8
0.5
1.0
1.5
2.0
2.5
3.0
47
60
73
0.59
0.75
0.91
1.11
0.98
1.09
88[b]/63[c]
79
87
0.8
0.8
0.8
2.0
2.0
2.0
85
1.41
1.43
–
86[d]
33[e]
0.3
0.3
0.3
0.3
0.3
2.0
2.0
2.0
2.0
2.0
30[f]
68[g]
95[h]
63[i]
39[j]
0.19
0.43
0.59
0.39
0.25
0.1[k]
3.0
55
0.68
[a] Unless otherwise mentioned, the isolated enzyme was used in MES
buffer (50 mm, pH 7.0). [b] Analytical yield. [c] Yield of the isolated
product. [d] A cell-free extract was used. [e] Whole E. coli cells (3.25 g)
were used. [f] The reaction was carried out at pH 5.5. [g] The reaction
was carried out at pH 6.5. [h] The reaction was carried out at pH 7.0.
[i] The reaction was carried out at pH 7.5. [j] The reaction was carried out
at pH 8.5. [k] aG1P was used.
In conclusion, an exceptionally efficient and selective
transglycosidase process has been developed. This process
should provide the basis for the production of aGG as an
industrial chemical.[21] We expect that it will also promote the
development of projected applications of this compound,
particularly as an ingredient for cosmetic formulations.
Experimental Section
Synthesis of aGG: Sucrose (0.3–2.0m) or aG1P (0.1m) was incubated
with glycerol (0.5–3.0m) and recombinant sucrose phosphorylase
from Leuconostoc mesenteroides[22] (3–80 UmLÀ1) in 2-(N-morpholi-
no)ethanesulfonic acid (MES) buffer (50 mm, pH 5.5–8.5) at 308C
(agitation rate: 550 rpm) for up to 72 h. The reaction time depended
on the quantity of the enzyme. Samples taken after appropriate time
intervals were inactivated by heating and centrifuged. The isolated
enzyme was purified according to a reported procedure.[10]
Analytical methods: The concentration of d-fructose was deter-
mined in an enzymatic assay by using recombinant mannitol
dehydrogenase from Pseudomonas fluorescens.[23] A colorimetric
glucose oxidase/peroxidase assay was used to determine the concen-
tration of d-glucose. Glucose oxidase from Aspergillus niger, horse-
radish peroxidase, and o-dianisidine were purchased from Sigma. The
quantity of inorganic phosphate present was determined colorimetri-
cally at 850 nm.[24] HPLC analysis was performed on an aminex HPX-
87C column (Bio-Rad) at 858C and at a constant flow rate of
0.6 mLminÀ1 with deionized water as the mobile phase. The concen-
tration of aGG was determined directly by HPLC or indirectly from
the difference in the concentrations of d-fructose (or phosphate) and
d-glucose.
Figure 2. Synthesis of aGG under optimized reaction conditions: 0.8m
sucrose, 2.0m glycerol, 20 UmLÀ1 enzyme, pH 7.0. d-fructose,
*
~
d-glucose, ! aGG, sucrose.
*
workup (see Experimental Section). The chemical structure
[
]
*
of the product was confirmed by NMR spectroscopy.
1
[*] Assignment of the chemical shifts in aGG: H NMR: glucose: 1-H,
5.21; 2-H, 3.66; 3-H, 3.84; 4-H, 3.52; 5-H, 3.93; 6-Ha, 3.95; 6-Hb,
3.83 ppm; glycerol: 1’-Ha, 3.80; 1’-Hb, 3.75; 2’-H, 3.91 ppm.
13C NMR: glucose: C1, 98.2; C2, 71.9; C3, 73.3; C4, 69.9; C5, 72.4;
C6, 61.0 ppm; glycerol: C1’, 63.8; C2’, 79.2 ppm.[8a] Chemical shifts
are referenced to external acetone: d(1H)=2.22 ppm, d(13C)=
31.5 ppm.
Isolation of aGG: The product mixture (16.2 g aGG in 90 mL)
was loaded onto an XK 50/60 column. The column material (1 L)
consisted of a 1:1 mixture of activated charcoal Norit (type Norit SX
ultra, Sigma) and calcined Celite 501 (Sigma). A four-step elution
gradient was used: The products were eluted with deionized water
10088 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 10086 –10089

Angewandte
Chemie
(4 L) followed by 2% ethanol/water (4 L), 15% ethanol/water (2 L),
and 25% ethanol/water (1 L) at a constant flow rate of 20 mLminÀ1
The fractions in 2% ethanol, which contained aGG (10.2 g,
40 mmol), were pooled, concentrated under vacuum, lyophilized,
and stored at À218C.
[8] a) R. Suhr, O. Scheel, J. Thiem, J. Carbohydr. Chem. 1998, 17,
.
937 – 968; b) N. Ohta, K. Achiwa, Chem. Pharm. Bull. 1991, 39,
1337 – 1339; c) A. Roder, E. Hoffmann, M. Hagemann, G. Berg,
FEMS Microbiol. Lett. 2005, 243, 219 – 226.
[9] a) C. Goedl, A. Schwarz, M. Mueller, L. Brecker, B. Nidetzky,
Carbohydr. Res. 2008, 343, 2032 – 2040; b) J. J. Mieyal, R. H.
Abeles, in The Enzymes, Vol. 7, 3rd ed., Academic Press, New
York, 1972, pp. 515 – 532.
Received: July 22, 2008
Published online: November 21, 2008
[10] A. Schwarz, B. Nidetzky, FEBS Lett. 2006, 580, 3905 – 3910.
[11] D. H. Crout, G. Vic, Curr. Opin. Chem. Biol. 1998, 2, 98 – 111.
[12] M. Scigelova, S. Singh, D. H. G. Crout, J. Mol. Catal. B 1999, 6,
483 – 494.
[13] H. Nakano, T. Kiso, K. Okamoto, T. Tomita, M. B. Manan, S.
Kitahata, J. Biosci. Bioeng. 2003, 95, 583 – 588.
Keywords: enzymatic synthesis · glycosides · kinetics ·
regioselectivity · transglucosylation
.
[1] a) H. Santos, P. Lamosa, T. Q. Faria, N. Borges, C. Neves in
Physiology and Biochemistry of Extremophiles (Eds.: C. Gerday,
N. Glansdorff), ASM Press, Washington, DC, 2007, pp. 86 – 103;
b) N. Empadinhas, M. S. da Costa, Int. Microbiol. 2006, 9, 199 –
206; c) M. S. da Costa, H. Santos, E. A. Galinski, Adv. Biochem.
Eng./Biotechnol. 1998, 61, 117 – 153.
[2] a) G. Lentzen, T. Schwarz, Appl. Microbiol. Biotechnol. 2006, 72,
623 – 634; b) “Orally used compatible solute containing agents”:
T. Schwarz, G. Lentzen, J. Krutmann, Germany, WO
2006097263, 2006.
[14] F. A. Shaikh, S. G. Withers, Biochem. Cell Biol. 2008, 86, 169 –
177.
[15] a) M. Faijes, A. Planas, Carbohydr. Res. 2007, 342, 1581 – 1594;
b) G. Perugino, A. Trincone, M. Rossi, M. Moracci, Trends
Biotechnol. 2004, 22, 31 – 37; c) G. J. Davies, S. J. Charnock, B.
Henrissat, Trends Glycosci. Glycotechnol. 2001, 13, 105 – 120.
[16] S. Kitao, H. Sekine, Biosci. Biotechnol. Biochem. 1992, 56, 2011 –
2014.
[17] O. Mirza, L. K. Skov, D. Sprogoe, L. A. van den Broek, G.
Beldman, J. S. Kastrup, M. Gajhede, J. Biol. Chem. 2006, 281,
35576 – 35584.
[18] a) E. J. Vandamme, J. van Loo, L. Machtelinckx, A. de Laporte,
in Adv. Appl. Microbiol., Vol. 32, Academic Press, New York,
1987, pp. 163 – 201; b) A. Schwarz, L. Brecker, B. Nidetzky,
Biochem. J. 2007, 403, 441 – 449.
[19] S. J. Williams, S. G. Withers, Carbohydr. Res. 2000, 327, 27 – 46.
[20] S. M. Hancock, M. D. Vaughan, S. G. Withers, Curr. Opin. Chem.
Biol. 2006, 10, 509 – 519.
[3] D. K. Hincha, M. Hagemann, Biochem. J. 2004, 383, 277 – 283.
[4] S. Fulda, J. Huckauf, A. Schoor, M. Hagemann, J. Plant Physiol.
1999, 154, 240 – 249.
[5] a) “a-d-Glucopyranosyl glycerol derivatives as antiallergic
agents, health foods, and cosmetics”: K. Yoshida, A. Takenaka,
T. Nitta, M. Iki, Japan, JP 2007137862, 2007; b) “Cosmetic
preparations containing glucosyl glycerides and one or more
acrylamidomethyl propylsulphonic acid polymers”: U. Breiten-
bach, V. Kallmayer, T. Raschke, C. Scherner, W. Siefken, S.
Viala, Germany, WO 2006122669, 2006; c) “Cosmetic prepara-
tions with an effective amount of glycosylglycerides as skin
moisturizers”: J. Thiem, O. Scheel, G. Schneider, Germany, EP
770378, 1997.
[21] “Preparation of 2-O-glyceryl-a-d-glucopyranoside from a glu-
cosyl donor and a glucosyl acceptor”: C. Goedl, T. Sawangwan,
B. Nidetzky, M. Mueller, Austria, WO 2008034158, 2008.
[22] C. Goedl, A. Schwarz, A. Minani, B. Nidetzky, J. Biotechnol.
2007, 129, 77 – 86.
[6] F. Takenaka, H. Uchiyama, Biosci. Biotechnol. Biochem. 2000,
64, 1821 – 1826.
[23] M. Slatner, B. Nidetzky, K. D. Kulbe, Biochemistry 1999, 38,
10489 – 10498.
[7] a) “Antitumor agents containing a-d-glucopyranosylglycerols
and food and cosmetics containing them”: T. Nitta, A. Takenaka,
M. Iki, E. Matsumura, M. Sakaguchi, JP 2007262023, 2007; b) D.
Colombo, F. Compostella, F. Ronchetti, A. Scala, L. Toma, H.
Tokuda, H. Nishino, Eur. J. Med. Chem. 2000, 35, 1109 – 1113.
[24] S. Saheki, A. Takeda, T. Shimazu, Anal. Biochem. 1985, 148,
277 – 281.
[25] Y. W. Kim, H. Chen, S. G. Withers, Carbohydr. Res. 2005, 340,
2735 – 2741.
Angew. Chem. Int. Ed. 2008, 47, 10086 –10089
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 10089