Enzymatic synthesis of octyl glucoside catalyzed by almond β-glucosidase in organic media

Article abstract of DOI:10.1139/v02-081

The synthesis of n-octyl-β-D-glucopyranoside can be performed by direct condensation of 1-octanol and glucose, catalyzed by immobilized almond β-glucosidase, using solid glucose suspended in 1-octanol as a reaction media. Both the rate of reaction and the conversion could be enhanced by using acetonitrile and N,N-dimethylformamide (DMF) as co-solvents, the latter giving the best results. The rate of reaction was dependent on the concentration of DMF and on the initial water activity (a_w), with higher water activity fostering faster reactions. The rate increased with DMF concentration, up to 20% DMF, but diminished rapidly at higher concentrations. Product concentration could be increased from 40 to 100 mM by going from 0 to 20% DMF; however, it was not sensitive to the initial water activity.

Full text of DOI:10.1139/v02-081

653
Enzymatic synthesis of octyl glucoside catalyzed
by almond -glucosidase in organic media
Amélie Ducret, Jean-François Carrière, Michael Trani, and Robert Lortie
Abstract: The synthesis of n-octyl-b-D-glucopyranoside can be performed by direct condensation of 1-octanol and glu-
cose, catalyzed by immobilized almond b-glucosidase, using solid glucose suspended in 1-octanol as a reaction media.
Both the rate of reaction and the conversion could be enhanced by using acetonitrile and N,N-dimethylformamide
(DMF) as co-solvents, the latter giving the best results. The rate of reaction was dependent on the concentration of
DMF and on the initial water activity (aw), with higher water activity fostering faster reactions. The rate increased with
DMF concentration, up to 20% DMF, but diminished rapidly at higher concentrations. Product concentration could be
increased from 40 to 100 mM by going from 0 to 20% DMF; however, it was not sensitive to the initial water activity.
Key words: b-glucosidase, alkyl glycosides, reverse hydrolysis, water activity, organic solvent.
Résumé : La synthèse du n-octyl-b-D-glucopyranoside a été effectuée par condensation du octane-1-ol et du glucose,
catalysée par la b-glucosidase d’amande immobilisée, en suspendant le glucose solide dans l’octane-1-ol. La vitesse de
réaction de même que le taux de conversion ont été améliorés par addition de co-solvants tels que l’acétonitrile et le
N,N-diméthylformamide (DMF), ce dernier permettant d’obtenir de meilleurs résultats. La vitesse de réaction dépendait
de la concentration en DMF et de l’activité de l’eau initiale, des activités élevées favorisant des réactions plus rapides.
La vitesse de réaction augmentait avec la concentration en DMF, jusqu’à 20% de DMF, mais diminuait rapidement à
des concentrations supérieurs. On a pu augmenter la concentration en produit, de 40 à 100 mM en passant de 0 à 20%
de DMF; cependant, elle n’était pas fonction de l’activité de l’eau.
Mots clés : b-glucosidase, glycosides d’alkyles, hydrolyse inversée, activité de l’eau, solvant organique.
Ducret et al.
656
Introduction
properties and low toxicity (11), and they have numerous
applications in areas such as cosmetics, pharmaceuticals,
food, and detergents. As a good alternative to chemical syn-
theses, which require protection and deprotection of the
hydroxyl groups, the use of glycosidases in the enzymatic
synthesis of such compounds was emphasized since it per-
mits the regio- and stereoselectivity to be obtained in one
step under mild reaction conditions. Enzymatic synthesis of
alkyl glycosides can be achieved by either a kinetically
(transglycosylation) or a thermodynamically (condensation)
controlled approach (12). In spite of numerous studies per-
formed on the subject (13), only poor yields have been ob-
tained, mostly because glycosidases lose activity at low
water activity (14, 15).
In the present work, we report the effect of water and co-
solvent on the activity of almond b-glucosidase in the syn-
thesis of octyl-b-D-glucoside by direct condensation of
glucose and 1-octanol. Almond b-glucosidase was selected
because it is readily available, relatively inexpensive, and
shows no inhibition by glucose when compared with various
microbial glycosidases (16).
The use of enzymes in organic synthesis is now widely
accepted and documented (1, 2). Simple enzymes like hy-
drolases have also found applications in synthesis (3). They
are especially interesting because they do not require co-
factors and many of them display good stability. Their use
has expanded with the recognition that they can catalyze the
reverse reaction, condensation, if the thermodynamic condi-
tions are favourable, by, for instance, lowering the water ac-
tivity (aw) by working in organic solvents. This fact has been
known for some time (4, 5), but became more prominent in
the last 20 years or so (6, 7), and some rationalizations were
made about the choice of solvent and the role of water (8, 9).
This approach can be used in the preparation of alkyl
glycosides. These non-ionic surfactants obtained from re-
newable feedstock are stable under alkaline conditions (10)
and completely biodegradable. They show antimicrobial
Received 8 November 2001. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 30 May 2002.
Dedicated to Professor J. Bryan Jones on the occasion of his
65th birthday.
Materials and methods
A. Ducret, J.-F. Carrière, M. Trani, and R. Lortie.1
Microbial and Enzyme Technology Group, Bioprocess Sector,
Biotechnology Research Institute, National Research Council,
6100 Royalmount Ave., Montreal, QC H4P 2R2, Canada.
Materials
Almond b-glucosidase (2.8 U mg–1), n-octyl-b-D-gluco-
pyranoside, p-nitrophenyl-b-D-glucopyranoside (pNPG), and
2-methyl-2-butanol were purchased from Sigma (St. Louis,
1Corresponding author (e-mail: robert.lortie@nrc.ca).
Can. J. Chem. 80: 653–656 (2002)
DOI: 10.1139/V02-081
© 2002 NRC Canada

654
Can. J. Chem. Vol. 80, 2002
Fig. 1. Synthesis of octyl-b-D-glucoside with aw = 0.90 (300 mg
D-glucose, 2 mL 1-octanol–DMF, 207 mg immobilized enzyme).
MO, U.S.A.). Glucose, Amberlite XAD-4 resin, sodium ace-
tate, and MgCl2 were purchased from BDH (Montreal, Que-
bec, Canada). The solvents used were always from the
highest grade available and were used without further purifi-
cation. Acetonitrile, 1-octanol, and the molecular sieves (3 Å)
were purchased from Fisher Scientific (Nepean, Ontario,
Canada). N,N-dimethylformamide (DMF), BaCl2, and
Mg(NO3)2 were purchased from Aldrich (Milwaukee, WI,
U.S.A.). Methanol was purchased from J.T. Baker
(Phillipsburgh, NJ, U.S.A.) and sulphuric acid from
Anachemia (Montreal). NaCl was purchased from ACP
Chemicals Inc. (Montreal). MilliQ+ water was used.
Immobilization of almond b-glucosidase on Amberlite
XAD-4
Almond b-glucosidase (483 mg) was dissolved in sodium
acetate buffer (2 mL, 50 mM, pH 5.0) and gently mixed with
XAD-4 resin (2 g), which was previously washed and dried
according to the procedure of Vulfson et al. (17). The mix-
ture was then dried under vacuum for 20 h. The immobilized
enzyme preparation contains 194.5 mg protein per g of dry
weight. The immobilization yield was measured by suspend-
ing the preparation (3.5 mg) in a buffer (50 mL) containing
pNPG (1.67 mM). Aliquots (0.75 mL) were taken, diluted
with glycine buffer (1.0 mL, 0.4 M, pH 10.8), and the
amount of p-nitrophenol released was measured at 430 nm.
reaction time (h)
octanol–DMF) were equilibrated separately in chambers
containing saturated salt solutions to preset the water activ-
ity (BaCl2, aw = 0.90; NaCl, aw = 0.75; Mg(NO3)2, aw
=
0.53; MgCl2, aw = 0.33). Equilibration of the different reac-
tion media was followed by measuring their water content
by coulometric Karl Fischer titration using a Metrohm 684
KF coulometer (Herisau, Switzerland). After addition of the
enzyme in the reaction mixture, the vials were hermetically
closed and the reactions were run at 50°C. In the reaction
mixture, most of the glucose was present as a solid phase. It
was verified that all the octyl glucoside formed during the
reaction was soluble in the liquid phase. Aliquots of the liq-
uid phase were taken periodically through a Mininert® valve
(Supelco, Bellefonte, PA, U.S.A.) for HPLC analysis. The
initial reaction rates were determined by linear regression
over 4–7 points.
HPLC analysis
HPLC analyzes were performed using a Waters Millen-
nium32 liquid chromatography system equipped with a Wa-
ters 600E solvent delivery system,
a
Waters 717
autosampler, and a Waters 410 refractive index detector, all
purchased from Waters Scientific (Montreal).
The analysis of octyl glucoside was performed using a re-
versed-phase column CSC-Inertsil, 150 Å/ODS2, 5 m, 25 ×
0.46 cm (CSC, Montreal). Periodically, 12 L of reaction
sample were dissolved in acetonitrile (320 L) and 25
L
Solubility measurement
were injected. A mixture of acetonitrile–water (60:40, v:v)
was used as the mobile phase. The flow rate was maintained
constant at 1.0 mL min–1 and the column temperature at
25°C. The retention times of n-octyl-b-D-glucopyranoside
and 1-octanol were 3.53 and 9.08 min, respectively.
The solid glucose as well as each solvent were previously
equilibrated separately in chambers containing a saturated
NaCl solution to preset the aw to 0.75. Glucose
(0.30–0.35 g) was then added to 2 mL of a mixture of 1-
octanol–co-solvent. The hermetically closed vials containing
the mixtures were kept under constant agitation at 50°C for
several days and then transferred to an oven at the same tem-
perature. The liquid phase was filtered (0.45 m, Millipore,
Bedford, MA, U.S.A.) and diluted with methanol for HPLC
analysis. To prevent precipitation, all the equipment used
was previously heated to 50°C. A second determination was
performed by simply diluting the settled liquid phase in
methanol before HPLC analysis. Both methods gave identi-
cal results.
Glucose analysis was accomplished on an ion-exclusion
column ICSep ICE-ION-300, 8 m, 30 × 0.78 cm preceded
by the guard column ICSep ICE-GC-801 (Transition Tech-
nologies Inc., Toronto, Ontario, Canada). A sample of the
mixture (100 L) was dissolved in methanol (200 L) and
10 L were injected into the column. Sulphuric acid (0.0042
M) was used as the mobile phase. The flow rate was main-
tained constant at 0.5 mL min–1 and the column temperature
at 60°C; 10 L of each sample were injected. The retention
times of glucose, methanol, acetonitrile, 2-methyl-2-butanol,
DMF, and n-octyl-b-D-glucopyranoside were 13.1, 25.5,
28.2, 46.0, 62.2, and 80.5 min, respectively. The system was
washed periodically with the mobile phase to elute 1-octanol
(which has a retention time of several hours).
Results and discussion
The immobilized preparation and the enzyme in solution
both catalyzed the hydrolysis of pNPG at a rate of 2.1 and
1.9 mol per min per mg protein, respectively. The immobi-
lization yield is therefore around 100%. When the solid was
filtrated from the reaction, the rate dropped to 20% of that
observed without filtration. This indicates that in water only
20% of the enzyme desorbs from the resin, even with a very
Enzymatic synthesis
Before starting the reactions, the immobilized enzyme
(207 mg, corresponding to 40.3 mg of protein) and reaction
media (300 mg glucose and 2 mL of a mixture of 1-
© 2002 NRC Canada

Ducret et al.
655
Fig. 2. Effect of various co-solvents on the initial reaction rate (a) and on the maximum concentration of octyl glucoside (b) (300 mg
D-glucose, 2 mL 1-octanol–co-solvent, 207 mg immobilized enzyme, aw = 0.75).
low percentages of acetonitrile (15–30%) and then rapidly
Fig. 3. Effect of co-solvent on the glucose solubility at aw = 0.75.
diminish as the concentration of acetonitrile increases. When
acetonitrile and DMF were used as co-solvents in water for
the synthesis of 2-hydroxybenzyl-b-D-glucopyranoside cata-
lyzed by almond b-glucosidase, the effect was not as favour-
able (18). In this case, acetonitrile had a positive effect on
the product concentration between 80 and 95% (v/v), while
DMF had no real positive effect.
The addition of polar co-solvents increases the solubility
of glucose in the reaction medium (Fig. 3), which could be
responsible for the observed increase in the reaction rates
upon the addition of small amounts of the co-solvent. The
fact that the addition of 2-methyl-2-butanol and acetonitrile
has a similar effect on the solubility of glucose, but a very
different effect on the initial reaction rates indicates that an
increase in glucose solubility is not a determining factor.
It is possible that the low reaction rate in octanol is partly
caused by its low dielectric constant (8.1), which results in
higher electrostatic interactions between the polar and the
charged residues, thus decreasing the flexibility of the pro-
tein. Addition of co-solvents with higher dielectric constants
such as DMF (38.2) or acetonitrile (36.6) (or increasing the
water content) raises the dielectric constant, which leads to
lower electrostatic interactions and higher protein flexibility
and activity, whereas 2-methyl-2-butanol, with its low di-
electric constant (5.8) has little impact on the activity. This
is in agreement with previous observations for subtilisin sus-
pended in tetrahydrofuran (19) and for a-chymotrypsin sus-
pended in various solvents (20).
high liquid–solid ratio. In a much smaller volume of octanol,
the enzyme should remain adsorbed (even in the presence of
hydrophilic solvents).
A typical time course of the b-glucosidase-catalyzed con-
densation of glucose and octanol, at an initial aw = 0.90 and
various DMF concentrations, is presented in Fig. 1. As can
be seen, the presence of the solvent affects both the initial
rate and the equilibrium-product concentration. As shown in
Fig. 2, other solvents at aw = 0.75 have a different effect on
the rate and equilibrium. The presence of 2-methyl-2-
butanol impairs the synthesis by diminishing the maximum
concentration of octyl glucoside reached as well as the ini-
tial rates at high concentrations (>50%). Conversely, addi-
tion of DMF or acetonitrile to the media increases the
production of octyl glucoside — with a different maximum
for each solvent. A sharp optimum concentration of DMF in
the media (20%) led to the highest rate as well as the highest
concentration of octyl glucoside at equilibrium (0.053 mol
per h per mg protein and 91 mM, respectively). Identical
high values for the initial rates can be obtained with a broad
concentration range of acetonitrile (40–65%); high concen-
trations of octyl glucoside, however, are only obtained with
Since DMF fosters higher reaction rates and product con-
centrations at aw = 0.75, a study of its effect at various water
activities was performed. As can be seen in Fig. 4, the posi-
tive effect of DMF on the reaction rates observed at aw
=
0.75 is present at other water activities up to a concentration
of 20%, after which the rates drop dramatically. The pres-
ence of DMF, up to 20%, fosters higher octyl glucoside
equilibrium concentrations. For instance, at aw = 0.53, nearly
100 mM (~30 g L–1) of octyl glucoside is obtained at equi-
librium, which is greater than the amount obtained without
the addition of co-solvents (40–50 mM); the amount mea-
sured by Ljunger et al. (14) was the same. Furthermore, this
© 2002 NRC Canada

656
Can. J. Chem. Vol. 80, 2002
Fig. 4. Effect of water activity on the initial reaction rate (a) and effect of DMF on the equilibrium concentration of octyl-b-D-
glucopyranoside (b) (same conditions as Fig. 1).
4. J.H. Kastle and A.S. Loevenhart. Am. Chem. J. 24, 491
(1900).
5. E.A. Sym. Biochem. J. 30, 609 (1936).
result has been obtained with a reaction rate superior to the
highest rates observed without DMF. This high conversion
is probably caused by the higher solubility of octyl
glucoside in the presence of DMF (8). The solubility of glu-
cose does not have to be taken into account, since the solu-
tion is always saturated, due to the presence of an excess of
solid glucose, leading to small variations in the thermody-
namic activity of glucose. It is difficult to explain the drop
in conversion at concentrations higher than 20% DMF. We
believe that equilibrium was reached, since addition of fresh
catalyst does not change concentrations.
It should be noticed that the maximum concentration of
octyl glucoside obtained increases only with the percentage
of added DMF and, surprisingly, does not seem to be influ-
enced by aw (Fig. 4b). This is contrary to observations for 1-
octanol without addition of co-solvent (14) and for 1-hexanol
using various glycoside donors (21), where in both cases, the
lowering of aw induces higher octyl glucoside concentra-
tions. The synthesis of 2-hydroxybenzyl-b-D-glucopyrano-
side in an aqueous-organic media was also favoured by low
water activities (18). The presence of most of the glucose as
a solid phase, on which water molecules can adsorb, may
play a role, which should be investigated.
6. A.M. Klibanov. ChemTech, 354 (1986).
7. A.M. Klibanov. Acc. Chem. Res. 23, 114 (1990).
8. P.J. Halling. Enzyme Microbial Technol. 16, 178 (1994).
9. P.J. Halling. Trends Biotechnol. 7, 50 (1989).
10. K. Holmberg. In Novel surfactants: Preparation, applications,
and biodegradablity. Edited by K. Holmberg. Marcel Dekker,
N.Y. 1998. p. 333.
11. S. Matsumura, K. Imai, S. Yoshikawa, K. Kawada, and T.
Uchibori. J. Am. Oil Chem. Soc. 67, 996 (1990).
12. V. Kasche. Enzyme Microbial Technol. 8, 4 (1986).
13. F. van Rantwijk, M. Woudenberg-van Oosterom, and R.A.
Sheldon. J. Mol. Catal. B: Enzym. 6, 511 (1999).
14. G. Ljunger, P. Adlercreutz, and B. Mattiasson. Enzyme Micro-
bial Technol. 16, 751 (1994).
15. M. Garcia-Garibay, A. Lopez-Mungia, and E. Barzana.
Biotechnol. Bioeng. 70, 647 (2000).
16. A. Ducret, M. Trani, and R. Lortie. Biotechnol. Bioeng. 77,
752 (2002).
17. E.N. Vulfson, R. Patel, and B.A. Law. Biotech. Lett. 12, 397
(1990).
18. G. Vic, D. Thomas, and D.H.G. Crout. Enzyme Microbial
Technol. 20, 597 (1997).
19. R. Affleck, Z.F. Xu, V. Suzawa, K. Focht, D.S. Clark, and J.S.
Dordick. Proc. Natl. Acad. Sci. U.S.A. 89, 1100 (1992).
20. R. Affleck, C.A. Haynes, and D.S. Clark. Proc. Natl. Acad.
Sci. U.S.A., 89, 5167 (1992).
21. M. Andersson and P. Adlercreutz. J. Mol. Catal. B: Enzym. 14,
69 (2001).
References
1. J.B. Jones. Tetrahedron, 42, 3351 (1986).
2. G.M. Whitesides and C.H. Wong. Angew. Chem. Int. Ed.
Engl. 24, 617 (1985).
3. U.T. Bornscheuer and R.J. Kazlauskas. In Hydrolases in or-
ganic synthesis: Regio- and stereoselective biotransformations.
Wiley-VCH, Weinheim. 1999.
© 2002 NRC Canada