Regio- and stereoselective glucosylation of diols by sucrose phosphorylase using sucrose or glucose 1-phosphate as glucosyl donor

Article abstract of DOI:10.1016/j.molcatb.2010.08.009

Previously it has been shown that glycerol can be regioselectively glucosylated by sucrose phosphorylase from Leuconostoc mesenteroides to form 2-O-α-d-glucopyranosyl-glycerol (Goedl et al., Angew. Chem. Int. Ed. 47 (2008) 10086-10089). A series of compounds related to glycerol were investigated by us to determine the scope of the α-glucosylation reaction of sucrose phosphorylase. Both sucrose and glucose 1-phosphate (G1P) were applied as glucosyl donor. Mono-alcohols were not accepted as substrates but several 1,2-diols were readily glucosylated, proving that the vicinal diol unit is crucial for activity. The smallest substrate that was accepted for glucosylation appeared to be ethylene glycol, which was converted to the monoglucoside for 69%. Using high acceptor and donor concentrations (up to 2.5 M), sucrose or G1P hydrolysis (with H₂O being the 'acceptor') can be minimised. In the study cited above, a preference for glucosylation of glycerol on the 2-position has been observed. For 1,2-propanediol however, the regiochemistry appeared to be dependent on the configuration of the substrate. The (R)-enantiomer was preferentialy glucosylated on its 1-position (ratio 2.5:1), whereas the 2-glucoside is the major product for (S)-1,2-propanediol (1:4.1). d.e. _ps of 71-83% were observed with a preference for the (S)-enantiomer of the glucosides of 1,2-propanediol and 1,2-butanediol and the (R)-enantiomer of the glucoside of 3-methoxy-1,2-propanediol. This is the first example of stereoselective glucosylation of a non-natural substrate by sucrose phosphorylase. 3-Amino-1,2-propanediol, 3-chloro-1,2-propanediol, 1-thioglycerol and glyceraldehyde were not accepted as substrates. Generally, the glucoside yield is higher when sucrose is used as a donor rather than G1P, due to the fact that the released phosphate is a stronger inhibitor of the enzyme (in case of G1P) than the released fructose (in case of sucrose). Essentially the same results are obtained with sucrose phosphorylase from Bifidobacterium adolescentis.

Full text of DOI:10.1016/j.molcatb.2010.08.009

Journal of Molecular Catalysis B: Enzymatic 67 (2010) 219–224
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Regio- and stereoselective glucosylation of diols by sucrose phosphorylase
using sucrose or glucose 1-phosphate as glucosyl donor
Rokus Renirie, Aliaksei Pukin, Barend van Lagen, Maurice C.R. Franssen∗
Wageningen University, Lab of Organic Chemistry, Dreijenplein 8, 6703 HB Wageningen, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 April 2010
Received in revised form 6 August 2010
Accepted 12 August 2010
Available online 18 August 2010
Previously it has been shown that glycerol can be regioselectively glucosylated by sucrose phospho-
rylase from Leuconostoc mesenteroides to form 2-O-␣-d-glucopyranosyl-glycerol (Goedl et al., Angew.
Chem. Int. Ed. 47 (2008) 10086–10089). A series of compounds related to glycerol were investigated
by us to determine the scope of the ␣-glucosylation reaction of sucrose phosphorylase. Both sucrose
and glucose 1-phosphate (G1P) were applied as glucosyl donor. Mono-alcohols were not accepted as
substrates but several 1,2-diols were readily glucosylated, proving that the vicinal diol unit is crucial
for activity. The smallest substrate that was accepted for glucosylation appeared to be ethylene glycol,
which was converted to the monoglucoside for 69%. Using high acceptor and donor concentrations (up
to 2.5 M), sucrose or G1P hydrolysis (with H2O being the ‘acceptor’) can be minimised. In the study cited
above, a preference for glucosylation of glycerol on the 2-position has been observed. For 1,2-propanediol
however, the regiochemistry appeared to be dependent on the configuration of the substrate. The (R)-
enantiomer was preferentialy glucosylated on its 1-position (ratio 2.5:1), whereas the 2-glucoside is the
major product for (S)-1,2-propanediol (1:4.1). d.e.ps of 71–83% were observed with a preference for the
(S)-enantiomer of the glucosides of 1,2-propanediol and 1,2-butanediol and the (R)-enantiomer of the
glucoside of 3-methoxy-1,2-propanediol. This is the first example of stereoselective glucosylation of a
non-natural substrate by sucrose phosphorylase. 3-Amino-1,2-propanediol, 3-chloro-1,2-propanediol, 1-
thioglycerol and glyceraldehyde were not accepted as substrates. Generally, the glucoside yield is higher
when sucrose is used as a donor rather than G1P, due to the fact that the released phosphate is a stronger
inhibitor of the enzyme (in case of G1P) than the released fructose (in case of sucrose). Essentially the
same results are obtained with sucrose phosphorylase from Bifidobacterium adolescentis.
Keywords:
Glucosylation
Sucrose phosphorylase
Regioselectivity
Stereoselectivity
Sucrose
Glucose 1-phosphate
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
A class of enzymes that has been intensely studied in their ability
to catalyse glucosylation with glucose as the donor are glucosidases
Glucosylation of alcohols and phenols is an important reaction
in nature and industry. Glucosides, oligo- and polysaccharides are
used by nature e.g. for energy storage, for strengthening cells and
tissues, and for molecular recognition. In pharmaceutical indus-
try, glucosylation is used to increase the solubility of drugs. Last
but not least, regio- and stereoselectively formed glucosides are
used as intermediates for other (bio)chemicals. Enzymatic glucosy-
lation naturally occurs mostly by tailor-made glucosyltransferases,
with a different glucosyltransferase for each glucosyl acceptor [1].
These transferases usually use UDP-glucose as glucosyl donor and
therefore the equilibria of the reactions catalysed are far on the glu-
coside side. However, the expensive nature of this donor is one of
the major bottlenecks in industrial application of these enzymes.
[1–4]. The equilibrium of the reaction catalysed by these enzymes is
however on the side of hydrolysis of the glucoside, leading to often
low yields with these enzymes. The equilibrium can be influenced
by the use of glucose derivatives with good leaving groups, like
4-nitrophenol [4], but these substrates are rather expensive. Alter-
native approaches involve media with minimal water content [5–9]
or continuous removal of the product [10] and they sometimes give
reasonable yields, but still they have inherent limitations.
More recently, phosphorylases have been extensively studied
with respect to glucosylation, one of these being sucrose phos-
phorylase. The enzyme from Bifidobacterium adolescentis has been
shown to catalyse the regioselective glucosylation of several sug-
ars such as l-arabinose [11]. The natural reaction catalysed by this
enzyme is the phosphorolysis of sucrose, with the concomitant
formation of ␣-glucose 1-phosphate (G1P) and fructose:
Abbreviations: G1P, Glucose ␣-1-phosphate; SUC, Sucrose; Fru, Fructose;
Glc, Glucose; Pi, Phosphate; d.e.p, Diastereomeric excess of the product.
sucrose + Pi → G1P + d-fructose
(1)
∗
The reaction proceeds through a ␤-glucosyl intermediate, which
is formed upon reaction of the enzyme with sucrose, as shown by
Corresponding author. Tel.: +31 317 482976; fax: +31 317 484914.
E-mail address: maurice.franssen@wur.nl (M.C.R. Franssen).
1381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.08.009

220
R. Renirie et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 219–224
donor (see below), 10 mM MgCl2 and 25 U/mL sucrose phospho-
rylase. The highest practical glucosyl acceptor concentration was
used, namely: 2.5 M ethylene glycol, 2.5 M (R,S)-1,2-propanediol,
2.5 M (R)-1,2-propanediol, 2.5 M (S)-1,2-propanediol, 2.5 M 1,2-
butanediol, 2.5 M 3-methoxy-1,2-propanediol, 0.25 M 3-chloro-
1,2-propanediol, 2.5 M 1-thioglycerol, 0.25 M glyceraldehyde,
1.7 M 3-amino-1,2-propanediol. The 3-amino-1,2-propanediol
stock solution was brought to pH 7 with 37% HCl prior to use. Con-
versions were followed in time by HPLC. The pH of the mixtures was
checked after completion. In the case of G1P a maximum increase
to pH 7.0 was observed. Two standard controls for all reactions
were performed: (i) same conditions without enzyme; (ii) same
conditions without glucosyl acceptor. For product isolation with a
charcoal column, 1.0 M sucrose and 1.0 M 1,2-butanediol were used
(1 mL scale); this reaction was followed using HPLC.
Fig. 1. Potential glucosyl acceptors similar to glycerol used in this study.
X-ray crystallography studies [12]. If phosphate is absent, other
glucosyl acceptors can be glucosylated:
sucrose + acceptor → ␣-glucoside + d-fructose
(2)
This reaction occurs with net retention of the ␣-glucosidic bond.
The equilibrium of this reaction is between the extreme situations
of the glucosyl transferases and glucosidases; the exact equilibrium
depends on the nature of the glucosyl acceptor. Both sucrose and
G1P can be used to form the glucosyl–enzyme intermediate. In the
latter case phosphate is released rather than fructose:
2.3. HPLC analysis
An Alltech OA-1000 column plus a wide pore C4 Security
Guard (Phenomenex, Utrecht, The Netherlands) were used with
25 mM H2SO4 as eluent at a flow rate of 0.4 mL min−1, cou-
pled to a refractive index detector (Gilson M 131, Gilson France,
Villiers le Bel, France). Samples were diluted 10-fold in water
after which 20 ␮L was injected via a fixed volume loop. Cal-
ibration curves for phosphate, glucose, fructose, sucrose and
glucosyl acceptors were linear in the concentration range mea-
sured; the same holds for G1P after correction for the overlap
with the negative void peak. Concentrations of these com-
pounds were calculated using these curves; for the glucosides
the concentrations were calculated assuming that glucose produc-
tion + glucoside production = glucoside-1-phosphate consumption
(=phosphate production).
G1P + acceptor → ␣-glucoside + Pi
(3)
Recently, Nidetzky and co-workers [13] have shown that glyc-
erol can be regioselectively converted with high yield using
sucrose phosphorylase from Leuconostoc mesenteroides. 2-O-(␣-
d-glucopyranosyl)-sn-glycerol, a natural osmolyte, is formed and
is currently marketed under the trade name Glycoin® [14]. To
determine the scope and selectivity of sucrose phosphorylase, a
series of compounds related to glycerol were investigated by us
(Fig. 1), using the enzymes from L. mesenteroides and B. adolescentis.
The formation of glucosylated ethylene glycol (1), 1,2-propanediol
(2), 1,2-butanediol (3) and 3-methoxy-1,2-propanediol (4) is
described, using sucrose or G1P as the glucosyl donor. The gluco-
sylation products of 1,2-propanediol, 3-methoxy-1,2-propanediol
and 1,2-butanediol were isolated and analysed. The glucosylation
reactions of 1,2-propanediol, 1,2-butanediol and 3-methoxy-1,2-
propanediol show interesting regio- and stereoselectivity.
2.4. TLC analysis
Prior to product isolation reactions were checked by TLC analy-
sis to determine optimal eluent conditions, using various ratio’s
of H2O, methanol and ethyl acetate. Spots were colored with a
molybdate-cerium based reagent (42 g L−1 Mo7O24(NH4)6·4H2O,
3.6 g L−1 Ce(NH4)(SO4)4·2H2O, 6.2% (v/v) H2SO4).
2. Experimental
2.1. Materials
2.5. Product isolation
Sucrose phosphorylase from L. mesenteroides, sucrose, glucose
1-phosphate, glycerol, ethylene glycol, (R,S)-1,2-propanediol,
(R)-1,2-propanediol, (S)-1,2-propanediol, (R,S)-1,2-butanediol,
2.5.1. Silica column
Glucoside products from (R,S)-1,2-propanediol, (R)-1,2-
propanediol, (S)-1,2-propanediol, (R,S)-1,2-butanediol and
(R,S)-3-methoxy-1,2-propanediol obtained with G1P as glu-
cosyl donor were isolated using pre-packed silica cartridges
using the Isolera One flash chromatography system from Bio-
tage. This procedure led to the removal of most of the excess
glucosyl acceptor, some remaining G1P, the formed phosphate
and most of the formed glucose; glucoside regio- or stereoiso-
mers could not be separated in this way. A mixture of ethyl
acetate:methanol:water = 7:2:1 was used for elution. Fractions
were analysed by HPLC and pooled fractions were freeze dried and
taken up in D2O for NMR analysis.
(R,S)-3-methoxy-1,2-propanediol,
1-thioglycerol,
3-chloro-1,2-propanediol,
(R,S)-3-amino-1,2-propanediol,
(R)-3-amino-
1,2-propanediol, (S)-3-amino-1,2-propanediol, glyceraldehyde,
magnesium chloride, MES buffer, disodium hydrogenphos-
phate, hydrogen chloride were obtained from Sigma–Aldrich
(Zwijndrecht, The Netherlands). Sucrose phosphorylase from B.
adolescentis was produced and isolated as described before [11].
Methanol (Biosolve, Valkenswaard, The Netherlands) and ethy-
lacetate (Fischer Scientific, Loughborough, UK) were HPLC grade.
NANOpure water (18.3 Mꢀ cm) was used during all experiments
(Barnstead, Thermo Scientific). Pre-packed flash purification silica
cartridges (SNAP 10 g KP-SIL) were obtained from Biotage.
2.5.2. Charcoal column
Using equimolar concentrations of 1,2-butanediol and sucrose
(1.0 M) resulted in 87% sucrose consumption after 96 h reaction
time. The product mixture consisted of 85% glucosides (correspond-
ing to 750 mM) and 15% glucose (120 mM). The glucoside product
mix was isolated with a charcoal column based on the procedure
that Goedl et al. [13,14] used for the isolation of the glucoside of
glycerol. This method removed sucrose, formed fructose and glu-
cose and remaining 1,2-butanediol, but could not separate product
2.2. Enzymatic glucosylation
Typical enzymatic reactions (250 ␮L) were done using 100 mM
G1P or 100 mM sucrose. A stock solution of G1P was brought
to pH 6.6 using HCl without adding additional buffer, whereas
in the case of sucrose (100 mM) MES buffer pH 6.6 (50 mM)
was added. Incubations at T = 30 ◦C further contained glucosyl

R. Renirie et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 219–224
221
regio- or stereoisomers. A pre-packed silica cartridge (SNAP 50 g
KP-SIL) was emptied and filled with a 1:1 mixture of activated char-
coal Norit (type Norit SX Ultra, Sigma–Aldrich) and calcined Celite
503 filter aid (Sigma–Aldrich). Elution was done in 4 steps: 250 mL
H2O, 250 mL 2% EtOH, 140 mL 15% EtOH, 90 mL 25% EtOH. Fractions
were dried, taken up in water and analysed by HPLC as described
above. On a 1 mL scale and using a charcoal column we obtained
an isolated yield of 124 mg glucoside mix (49%) with a purity >90%.
The compounds eluted in the order: fructose (0–2% EtOH), glucose
(2% EtOH), sucrose (15% EtOH), glucoside isomers (15–25% EtOH),
1,2-butanediol (25% EtOH).
2.6. NMR analysis
Pooled fractions from the silica column or charcoal column were
combined, freeze dried and taken up in 600 ␮L D2O for NMR anal-
ysis. 1H, 13C (APT), COSY, HSQC and HMBC spectra were taken on
a Bruker Avance III 400 MHz NMR spectrometer. Typically about
50 mg of product (purity 75–99%) was isolated for each conversion
and analysed by NMR. For the purity analysis of the samples see the
Supplementary file.
2.6.1. (R)-2-Hydroxy-1-propyl ˛-glucopyranoside (11, the major
regio-isomer that is formed)
1H NMR (400 MHz, D2O) ı (ppm): 4.87 (d, J = 3.3 Hz, 1H, H-1),
4.00 (1H, CH3CH(OH)CH2OGlc), 3.80 (1H, H-6), 3.69 (1H, H-6ꢀ),
3.67 (1H, H-3), 3.65 (1H, CH3CH(OH)CH2OGlc), 3.62 (1H, H-5), 3.49
(1H, H-2), 3.35 (1H, H-4), 3.31 (1H, CH3CH(OH)CH2OGlc), 1.12 (3H,
CH3). 13C NMR (101 MHz, D2O) ı (ppm): 98.6 (C-1), 73.1 (C-3),
72.9 (CH3CH(OH)CH2OGlc), 71.8 (C-5), 71.5 (C-2), 69.6 (C-4), 66.6
(CH3CH(OH)CH2OGlc), 60.5 (C-6), 18.1 (CH3).
2.6.2. (R)-1-Hydroxy-2-propyl ˛-glucopyranoside (10, minor
regio-isomer formed)
1H NMR (400 MHz, D2O) ı (ppm): 5.02 (1H, H-1), 3.84 (1H,
CH3CH(OGlc)CH2OH), 3.84–3.65 (2H, H-6, H-6ꢀ), 3.72 (1H, H-
5), 3.65 (1H, H-3), 3.59 (1H, CH3CH(OGlc)CH2OH), 3.51 (1H,
CH3CH(OGlc)CH2OH), 3.47 (1H, H-2), 3.36 (1H, H-4), 1.17 (3H,
CH3). 13C NMR (101 MHz, D2O)
ı (ppm): 98.3 (C-1), 75.8
(CH3CH(OGlc)CH2OH), 73.1 (C-3), 72.0 (C-5), 71.6 (C-2), 69.7 (C-4),
64.7 (CH3CH(OGlc)CH2OH), 60.6 (C-6), 16.9 (CH3).
2.6.3. (S)-1-Hydroxy-2-propyl ˛-glucopyranoside (12, major
regio-isomer formed)
1H NMR (400 MHz, D2O) ı (ppm): 4.99 (1H, H-1), 3.84 (1H,
CH3CH(OGlc)CH2OH), 3.79 (1H, H-6), 3.74 (1H, H-5), 3.70 (1H, H-
6ꢀ), 3.68 (1H, H-3), 3.59 (2H, CH3CH(OGlc)CH2OH), 3.49 (1H, H-2),
3.36 (1H, H-4), 1.08 (3H, CH3). 13C NMR (101 MHz, D2O) ı (ppm):
95.6 (C-1), 73.2 (CH3CH(OGlc)CH2OH), 73.0 (C-3), 71.8 (C-5), 71.3
(C-2), 69.6 (C-4), 65.3 (CH3CH(OGlc)CH2OH), 60.5 (C-6), 14.4 (CH3).
2.6.4. (S)-2-Hydroxy-1-propyl ˛-glucopyranoside (13, minor
regio-isomer formed)
1H NMR (400 MHz, D2O) ı (ppm): 4.86 (d, J = 2.8 Hz, 1H, H-
1), 4.02 (1H, CH3CH(OH)CH2OGlc), 3.80–3.70 (2H, H-6, H-6ꢀ), 3.66
(1H, H-3), 3.55 (1H, CH3CH(OH)CH2OGlc), 3.65 (1H, H-5), 3.48 (1H,
H-2), 3.34 (1H, H-4), 3.45 (1H, CH3CH(OH)CH2OGlc), 1.13 (3H,
CH3). 13C NMR (101 MHz, D2O) ı (ppm): 98.1 (C-1), 73.1 (C-3),
72.4 (CH3CH(OH)CH2OGlc), 71.8 (C-5), 71.5 (C-2), 69.6 (C-4), 66.2
(CH3CH(OH)CH2OGlc), 60.6 (C-6), 18.1 (CH3).
2.6.5. (S)-1-Hydroxy-2-butyl ˛-glucopyranoside (16, major
regio- and stereoisomer formed)
1H NMR (400 MHz, D2O) ı (ppm): 4.99 (1H, H-1), 3.77 (1H,
H-6), 3.74 (1H, H-5), 3.68 (1H, H-6ꢀ), 3.67 (1H, H-3), 3.62 (2H,
CH3CH2CH(OGlc)CH2OH), 3.60 (1H, CH3CH2CH(OGlc)CH2OH), 3.48

222
R. Renirie et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 219–224
(1H, H-2), 3.34 (1H, H-4), 1.55 (2H, CH3CH2CH(OGlc)CH2OH),
0.86 (3H, CH3). 13C NMR (101 MHz, D2O) ı (ppm): 97.1 (C-1),
79.5 (CH3CH2CH(OGlc)CH2OH), 73.0 (C-3), 71.7 (C-5), 71.6 (C-
2), 69.6 (C-4), 63.2 (CH3CH2CH(OGlc)CH2OH), 60.6 (C-6), 22.4
(CH3CH2CH(OGlc)CH2OH), 8.8 (CH3).
(not shown). This could also be an explanation for the previously
higher observed yield of glucosylation of glycerol using sucrose as
glucosyl donor compared to G1P [13].
We also tested glucosylation of all eight diols mentioned in this
study with sucrose phosphorylase from B. adolescentis [11,12], giv-
ing essentially the same results during TLC and HPLC analysis.
2.6.6. (R)-1-Hydroxy-3-methoxy-2-propyl ˛-glucopyranoside
(20, major regio- and stereoisomer formed)
3.2. Isolation and NMR analysis of product structure, regio- and
stereoselectivity
1H NMR (400 MHz, D2O) ı (ppm): 5.06 (d, J = 3.6 Hz, 1H,
H-1), 3.85 (1H, CH3OCH2CH(OGlc)CH2OH), 3.78 (1H, H-6),
3.75 (1H, H-5), 3.69 (1H, H-6ꢀ), 3.67 (1H, H-3), 3.66 (2H,
CH3OCH2CH(OGlc)CH2OH), 3.58 (2H, CH3OCH2CH(OGlc)CH2OH),
3.47 (1H, H-2), 3.36 (1H, H-4), 3.33 (3H, CH3). 13C NMR (101 MHz,
D2O) ı (ppm): 97.6 (C-1), 76.4 (CH3OCH2CH(OGlc)CH2OH), 72.9
(C-3), 72.1 (C-5), 71.6 (C-2), 71.3 (CH3OCH2CH(OGlc)CH2OH), 69.6
(C-4), 61.5 (CH3OCH2CH(OGlc)CH2OH), 60.6 (C-6), 58.5 (CH3).
In Fig. 3 the possible products of all conversions in this study
are shown, plus product 22 that was found for the glucosylation
of glycerol [13]. The product of ethylene glycol was not analysed
by NMR; only one peak was observed during HPLC analysis so we
assume it is the monoglucoside 9.
For glucosylation of (R,S)-1,2-propanediol there is an overall
preference for the 2-position, as earlier observed for gluco-
sylation of glycerol [13], but the ratio 2-glucoside:1-glucoside
((10 + 12):(11 + 13)) is less marked (1.8:1). Interestingly, when
we analysed glucosylation of the individual enantiomers we
observed the 1-regioisomer to be the major product for the (R)-
1,2-propanediol (10:11 = 1:2.5), whereas the 2-regioisomer is the
major product for (S)-1,2-propanediol (12:13 = 4.1:1). Overall this
means that the stereoselectivity R:S is 1:6 (d.e.p = 71%), assuming
that in the racemate the preference of the enzyme for the 1- or
2-position of either of the enantiomers is the same as that for the
individual enantiomers.
Also for glucosylation of (R,S)-1,2-butanediol and (R,S)-
3-methoxy-1,2-propanediol there is an overall preference
for the 2-position. The ratio 2-glucoside (14 + 16):1-glucoside
(15 + 17) is 5.6:1 for (R,S)-1,2-butanediol (3). Likewise, the ratio
(18 + 20):(19 + 21) is 7.4:1 for (R,S)-3-methoxy-1,2-propanediol
(4). Both ratios are larger than for (R,S)-1,2-propanediol (2). For
(R,S)-1,2-butanediol (3) the stereoselectivity is 1:6.6 (d.e.p = 74%) in
benefit of one of the enantiomers. If the analogy with propanediol
can be applied, the preference would be for the S-enantiomer
as well. For (R,S)-3-methoxy-1,2-propanediol (4) the stereose-
lectivity is even higher (1:11, d.e.p = 83%). If the analogy with
(R,S)-1,2-propanediol can be applied, the preference would be
for the R-enantiomer (which has the same three-dimensional
structure as (S)-1,2-propanediol).
Very recently it was shown by Schwartz et al. [15] that also
the ␤-glucoside of glycerol can be formed enzymatically using a
thermostable cellobiose phosphorylase from Pyrococcus furiosus
and cellobiose as the glucosyl donor. It is conceivable that also
interesting enantioselectivity will be observed with this enzyme
in combination with the compounds studied by us.
3. Results and discussion
3.1. Conversion of various alcohols with G1P as glucosyl donor
First, sucrose phosphorylase was tested for its ability to glucosy-
late mono-alcohols. Methanol and ethanol were not glucosylated.
After this we tested various diols resembling glycerol. The enzy-
matic glucosylation of (R,S)-1,2-butanediol with G1P as glucosyl
donor was studied by HPLC. A clear formation of a glucoside prod-
uct peak at tR = 12.2 min is observed, which is absent in the control
without enzyme (see Fig. S1, Supplementary material). In addition,
phosphate release is observed, plus the formation of some glucose.
As observed before for the ‘synthesis reaction’ of phosphorylases
[1], water competes as a ‘glycosyl acceptor’, in this case forming
glucose. Consequently, in the absence of (R,S)-1,2-butanediol, glu-
cose formation is higher. While testing the mono-alcohols ethanol
and methanol a similar amount of G1P hydrolysis was observed as
in their absence, indication that the enzyme was not simply dena-
tured or inhibited. At these time scales no glucose formation as a
consequence of non-catalytic hydrolysis of G1P is observed (see Fig.
S1, Supplementary material).
For (R,S)-1,2-butanediol, 55% of the G1P is consumed; an equi-
librium is reached as can be seen in Fig. 2 (panel A), where the
conversion in time is depicted. Fig. 2 also shows the conversions
of ethylene glycol, (R,S)-1,2-propanediol and (R,S)-3-methoxy-1,2-
propanediol (Panels B–D).
Table
1 summarises the amounts of glucosylated prod-
ucts formed for all glycerol analogues that were investi-
gated. For (R,S)-1,2-propanediol, ethylene glycol and (R,S)-
3-methoxy-1,2-propanediol, >85% of G1P is consumed. The
rate of product formation increases in the order (R,S)-1,2-
propanediol → (R,S)-3-methoxy-1,2-propanediol → ethylene gly-
col → (R,S)-1,2-butanediol, but differences are not large. With
(R,S)-1,2-propanediol a relatively large amount of G1P hydrol-
ysis is observed. For (R,S)-3-amino-1,2-propanediol (5), (R,S)-
3-chloro-1,2-propanediol (6), (R,S)-1-thioglycerol (7) and (R,S)-
glyceraldehyde (8) no glucoside product formation was observed.
3.3. Conversion of 1,2-butanediol: G1P vs. sucrose as glucosyl
donor
Having identified four new substrates for sucrose phosphory-
lase, we subsequently compared the use of (cheap) sucrose and G1P
as glucosyl donors for this enzyme. As expected, all glucosides were
also formed using sucrose as a donor. Fig. 4 shows the glucosylation
of 1,2-butanediol (3) as a function of time.
A higher conversion of aglycon is obtained with sucrose as a
donor; in contrast to G1P (see Fig. 2) sucrose is almost completely
consumed, which is most likely due to a more favorable equilib-
rium. However, the disadvantage of using sucrose is the release
of fructose. This byproduct is more cumbersome to remove com-
pared to phosphate in the case of G1P, which is also reflected in
the HPLC chromatograms (see Fig. S2, Supplementary material).
Thus for product identification via NMR we used isolated glucosides
made with G1P as a donor. Obviously, for industrial application
sucrose is much cheaper, but it might be conceivable that when
For (R,S)-1-thioglycerol
a product was observed which was
also formed in the absence of enzyme, presumably due
to non-enzymatic disulfide formation. The presence of (R,S)-
glyceraldehyde decreased glucose formation by 50%, indicating
binding in the active site without conversion. For 3-amino-1,2-
propanediol both stereoisomers were also tested separately, both
showing no conversion.
During the conversions a decrease in rate of product formation
in time is observed for all four products. To test if this is due to
inhibition by released phosphate, the experiment with ethylene
glycol was repeated with an initial phosphate concentration of
300 mM. The result was immediate inhibition for more than 95%

R. Renirie et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 219–224
223
Fig. 2. Glucosylation of (R,S)-1,2-butanediol (A), ethylene glycol (B), (R,S)-1,2-propanediol (C), and (R,S)-3-methoxy-1,2-propanediol (D) by sucrose phosphorylase from L.
mesenteroides as a function of time with G1P as glucosyl donor.
Fig. 3. Possible mono-glucosides of ethylene glycol (9), 1,2-propanediol (10–13), 1,2-butanediol (14–17) and 3-methoxy-1,2-propanediol (18–21) plus the observed single
glucosylation product of glycerol (22) [13]. (R)-Glucosyl acceptor units are indicated in blue and (S)-glucosyl acceptor units in green. (For the mentioned colors in this legend
the reader is referred to the web version of this article).

224
R. Renirie et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 219–224
conceivable that this type of stereoselectivity will also be found for
other phosphorylases converting glycerol-like compounds.
Note added in proof
The Nidetzki group has also very recently published on the
sucrose phosphorylase mediated glucosylation of 1,2-propanediol,
see [16].
Acknowledgements
This research was performed as part of the Program Integra-
tion of Biosynthesis & Organic Synthesis (IBOS) of ACTS-NWO (The
Netherlands).
Prof. Dr. Ron Wever is kindly thanked for the use of HPLC equip-
ment. Mr. Louis Hartog, Dr. Teunie van Herk, Dr. Teris van Beek
and Mr. Elbert van der Klift are acknowledged for their help during
HPLC measurements. Dr. Ben van den Broek and Dr. Gerrit Beldman
are kindly thanked for their generous gift of sucrose phosphorylase
from B. adolescentis.
Fig. 4. Glucosylation of 1,2-butanediol (3) by sucrose phosphorylase from L. mesen-
teroides as a function of time with sucrose as glucosyl donor.
G1P is generated enzymatically this route become cost-effective as
well. We have preliminary results that by starting from glycogen
and using a cascade of glycogen phosphorylase and sucrose phos-
phorylase it is possible to form the glucoside of ethylene glycol (not
shown).
With both glucosyl donors no side product formation apart from
glucose was observed; we anticipated some formation of mal-
tose, as a consequence of ␣-glucosylation of the formed glucose;
however, at the retention time of maltose (8.40 min) no peak was
observed in any of our experiments.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.molcatb.2010.08.009.
References
[1] C.A.G.M. Weijers, M.C.R. Franssen, G.M. Visser, Biotechnol. Adv. 26 (2008)
436–456.
[2] A. Millqvist-Fureby, I.S. Gill, E.N. Vulfson, Biotechnol. Bioeng. 60 (1998)
190–196.
4. Conclusions
[3] S. Takayama, G.J. McGarvey, C.H. Wong, Chem. Soc. Rev. 26 (1997) 407–415.
[4] F. van Rantwijk, M. Woudenberg-van Oosterom, R.A. Sheldon, J. Mol. Catal. B:
Enzyme 6 (1999) 511–532.
[5] Z. Chahid, D. Montet, M. Pina, J. Graille, Biotechnol. Lett. 14 (1992) 281–284.
[6] B.M. de Roode, S.W.P.G. Peters, M.C.R. Franssen, A. van der Padt, R.M. Boom,
Biotechnol. Bioeng. 72 (2001) 568–572.
[7] M. Gelo-Pujic, E. Guibe-Jampel, A. Loupy, A. Trincone, J. Chem. Soc. Perkin Trans.
1 7 (1997) 1001–1002.
[8] G. Ljunger, P. Adlercreutz, B. Mattiasson, Enzyme Microb. Technol. 16 (1994)
751–755.
[9] G. Vic, D. Thomas, D.H.G. Crout, Enzyme Microb. Technol. 20 (1997) 597–603.
[10] B.M. de Roode, M.C.R. Franssen, A. Van der Padt, R.M. Boom, Biotechnol. Prog.
19 (2003) 1391–1402.
[11] L.A.M. van den Broek, E.L. van Boxtel, R.P. Kievit, R. Verhoef, G. Beldman, A.J.G.
Voragen, Appl. Microbiol. Biotechnol. 65 (2004) 219–227.
[12] O. Mirza, L. Skov, D. Sprogoe, L.A.M. Van den Broek, G. Beldman, J.S. Kastrup, M.
Gajhede, J. Biol. Chem. 281 (2006) 35576–35584.
[13] C. Goedl, T. Sawangwan, M. Mueller, A. Schwartz, B. Nidetzky, Angew. Chem.
Int. Ed. 47 (2008) 10086–10089.
[14] C. Goedl, T. Sawangwan, B. Nidetzky, M. Mueller, Patent WO 2008/034158
(2008).
Four new compounds resembling glycerol were found to be
glucosylated by sucrose phosphorylase from L. mesenteroides
and B. adolescentis, namely ethylene glycol, 1,2-propanediol, 1,2-
butanediol and 3-methoxy-1,2-propanediol. Since mono-alcohols
are not converted, the minimal unit necessary for these compounds
to be accepted as substrates appears to be ethylene glycol. Both ␣-
glucose 1-phosphate and sucrose can be used as the glucosyl donor.
If one terminal hydroxyl group of glycerol is substituted, regiose-
lective ␣-glucosylation on the 2-position (overall) is observed for
the four compounds mentioned above, whereas for other com-
pounds no glucosylation is observed. We show that glucosylation
of the individual enantiomers of 1,2-propanediol results in a dif-
ferent regioselectivity for each enantiomer; the 1-regioisomer
(2-hydroxy-1-propyl ␣-glucopyranoside, 11) is the major product
for the reaction of (R)-1,2-propanediol, whereas the 2-regioisomer
(1-hydroxy-2-propyl ␣-glucopyranoside, 14) is the major product
for (S)-1,2-propanediol. For the compounds tested also enantiose-
lectivity is observed with ee’s up to 83%. This is the first time that
this stereoselectivity is observed for sucrose phosphorylase. It is
[15] A. Schwartz, M.S. Thomson, B. Nidetzky, Biotechnol. Bioeng. 103 (2009)
865–872.
[16] C. Luley-Goedl, T. Sawangwan, L. Brecker, P. Wildberger, B. Nidetzki, Carbohyd.
Res. 345 (2010) 1736–1740.