Appearance and distribution of regioisomers in metallo- and serine-protease-catalysed acylation of sucrose in N,N-dimethylformamide

Article abstract of DOI:10.1016/j.molcatb.2014.04.017

The appearance and distribution of monoester regioisomers were investigated in the virtually irreversible acylation of sucrose with the enol ester, vinyl laurate, as acyl donor catalysed by serine proteases and a metalloprotease in the hydrophilic, aprotic solvent N,N-dimethylformamide. Sucrose laurate was obtained in yields from 12 to 53% after 48 h under different catalytic conditions. The serine protease ALP-901, derived from a Streptomyces sp., produced the highest yield at this reaction time, while reaction with the zinc-protease thermolysin achieved the overall highest yield (63%) after 6 h, with only monoesters synthesised. The total conversion of sucrose after 48 h ranged from 19 to 96%. The highest degree of conversion was observed in the reaction with thermolysin, while the reactions without protein and with ALP-901 resulted in 82% and 66% sucrose conversion, respectively. 2-O-Lauroyl sucrose was the most abundant monoester regioisomer synthesised and the highest concentration observed was 23.7 mM after 24 h in the thermolysin-catalysed reaction. The highest concentration of 2-O-lauroyl sucrose detected in the reaction catalysed by ALP-901 was 19.0 mM, while it was 17.0 mM the reaction without protein, both after 48 h. The detected appearance of the sucrose laurate regioisomers largely corresponded to the apparent rates of formation, and 2-O-lauroyl sucrose was among the first regioisomers to appear in all reactions. The observed sucrose laurate regioisomeric distribution after 48 h (2:3:4:6:1′:3′) was 72:5:2:1:7:14 in the reaction catalysed by ALP-901, and 74:5:2:1:7:13 in the reaction without protein. In the reaction catalysed by thermolysin the distribution was 71:5:2:-:9:13 after 6 h and 86:8:-:-:4:3 after 48 h of reaction. The esterification of sucrose with vinyl laurate without protein in the reaction mixture appeared to be catalysed in the presence of aluminosilicate molecular sieves. Non-catalytic protein in the reaction medium seemed to lower the catalytic activity of the molecular sieves.

Full text of DOI:10.1016/j.molcatb.2014.04.017

Journal of Molecular Catalysis B: Enzymatic 106 (2014) 26–31
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Appearance and distribution of regioisomers in metallo- and
serine-protease-catalysed acylation of sucrose in
N,N-dimethylformamide
Aleksander Liea, Anne S. Meyerb, Lars Haastrup Pedersena,∗
a Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Sohngaardsholmsvej 49, 9000 Aalborg, Denmark
b Department of Chemical and Biochemical Engineering, Technical University of Denmark, Søltofts Plads, Building 227, 2800 Kgs. Lyngby, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
The appearance and distribution of monoester regioisomers were investigated in the virtually irreversible
acylation of sucrose with the enol ester, vinyl laurate, as acyl donor catalysed by serine proteases
and a metalloprotease in the hydrophilic, aprotic solvent N,N-dimethylformamide. Sucrose laurate was
obtained in yields from 12 to 53% after 48 h under different catalytic conditions. The serine protease
ALP-901, derived from a Streptomyces sp., produced the highest yield at this reaction time, while reac-
tion with the zinc-protease thermolysin achieved the overall highest yield (63%) after 6 h, with only
monoesters synthesised. The total conversion of sucrose after 48 h ranged from 19 to 96%. The highest
degree of conversion was observed in the reaction with thermolysin, while the reactions without pro-
tein and with ALP-901 resulted in 82% and 66% sucrose conversion, respectively. 2-O-Lauroyl sucrose
was the most abundant monoester regioisomer synthesised and the highest concentration observed
was 23.7 mM after 24 h in the thermolysin-catalysed reaction. The highest concentration of 2-O-lauroyl
sucrose detected in the reaction catalysed by ALP-901 was 19.0 mM, while it was 17.0 mM the reaction
without protein, both after 48 h. The detected appearance of the sucrose laurate regioisomers largely
corresponded to the apparent rates of formation, and 2-O-lauroyl sucrose was among the first regioi-
somers to appear in all reactions. The observed sucrose laurate regioisomeric distribution after 48 h
(2:3:4:6:1ꢀ:3ꢀ) was 72:5:2:1:7:14 in the reaction catalysed by ALP-901, and 74:5:2:1:7:13 in the reaction
without protein. In the reaction catalysed by thermolysin the distribution was 71:5:2:–:9:13 after 6 h
and 86:8:–:–:4:3 after 48 h of reaction. The esterification of sucrose with vinyl laurate without protein
in the reaction mixture appeared to be catalysed in the presence of aluminosilicate molecular sieves.
Non-catalytic protein in the reaction medium seemed to lower the catalytic activity of the molecular
sieves.
Received 3 February 2014
Received in revised form 11 April 2014
Accepted 23 April 2014
Available online 2 May 2014
Keywords:
Sucrose fatty acid ester
Regioisomeric distribution
Protease-catalysed acylation
Hydrophilic solvent
Non-enzymatic catalysis
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
product properties [2,5,6]. Biocatalysis has been shown as a strong
alternative to conventional chemical synthesis, offering improved
Sucrose fatty acid esters are widely used in the food, cos-
metics, pharmaceutical and detergent industries, as they can be
produced with a wide range of emulsifying, dispersing and solu-
bilising properties [1,2]. These non-ionic surfactants are non-toxic
and biodegradable, and some of them have shown antimicrobial
effects [1,3,4]. Currently sucrose alkanoates are synthesised by con-
ventional chemical processes at high temperatures, resulting in
low regiospecificity and side reactions leading to alterations of
activity and selectivity while not needing the protection and depro-
tection steps often required in chiral and regioselective organic
synthesis. In particular, lipases and proteases have been applied to
the synthesis of sucrose alkanoates in organic solvents [7]. Lipases
have been shown to catalyse esterification by reversed hydroly-
sis reactions and exhibit selectivity towards the primary hydroxyl
groups of sucrose (6, 1ꢀ and 6ꢀ) [8,9], while proteases catalyse trans-
esterification reactions and exhibit selectivity towards primary or
secondary hydroxyl groups in sucrose [10,11]. Hydrophilic, aprotic
solvents, including N,N-dimethylformamide (DMF) and dimethyl-
sulfoxide (DMSO) are excellent solvents for carbohydrates but
act as protein denaturants affecting enzyme activity and stabil-
ity. Proteases have shown improved activity and selectivity with
∗
Corresponding author. Tel.: +45 9940 8497; fax: +45 9814 1808.
E-mail addresses: alie@bio.aau.dk (A. Lie), am@kt.dtu.dk (A.S. Meyer),
lhp@bio.aau.dk (L.H. Pedersen).
http://dx.doi.org/10.1016/j.molcatb.2014.04.017
1381-1177/© 2014 Elsevier B.V. All rights reserved.

A. Lie et al. / Journal of Molecular Catalysis B: Enzymatic 106 (2014) 26–31
27
System peak
& sucrose
DMSO as co-solvent [12,13], while lipases have shown sensitiv-
ity to hydrophilic solvents, losing activity and exhibiting complete
inactivation at DMSO concentrations of 20–30% and above [14,15].
Subtilisin has shown selectivity towards 1ꢀ-OH and 6-OH of sucrose
in anhydrous DMF as well as in mixtures of buffer and DMSO
[16–18], while the metalloprotease thermolysin catalysed the acy-
lation of sucrose primarily at the 2-OH [19] and the alkaline
protease AL-89 catalysed the esterification of 2-OH and promoted
acyl migration to the 3-regioisomer of sucrose alkanoates [13].
The recent developments in methods for quantitative analysis of
sucrose laurate using RP-HPLC with charged aerosol detection
(CAD) [20] made more accurate studies of the esterification pro-
cesses possible, and in particular offered increased sensitivity in the
analysis of the distribution of regioisomers. In this work, we have
investigated catalytic activity and appearance and distribution of
monoester regioisomers in the irreversible acylation of sucrose
with the enol ester vinyl laurate as acyl donor catalysed by serine
proteases and a metalloprotease in the hydrophilic, aprotic solvent
DMF.
100
80
60
40
20
0
2
3’
1’
3
4
6
6’
25
4’
0
5
10
15
time (min)
20
30
Fig. 1. Chromatogram of reaction mixture from reaction catalysed by thermolysin
after 6 h. Peaks annotated with regioisomer substitution position. Retention times:
sucrose, 1.1; sucrose laurate regioisomers, 12.2, 14.5, 18.2, 20.5, 21.5, 22.3, 25.6,
29.5.
2.3. RP-HPLC analysis
Samples (10 ␮L injections) were analysed using a Dionex Ulti-
Mate 3000 system (Thermo Scientific) with Corona Veo CAD
charged aerosol detector (Thermo Scientific, N2 at 61.5 psi, neb-
uliser at 50 ◦C). Stationary phase was a Symmetry C18 column
(250 mm × 4.6 mm, 5 ␮m particle size, Waters) maintained at 45 ◦C,
and eluents consisted of mixtures of water and acetonitrile, both
with 0.05% (v/v) TFA, with flow rate 2.0 mL/min. The chromato-
graphic data were collected using Chromeleon software (version
7.2, Thermo Scientific).
Sucrose laurate regioisomer analysis was conducted using
isocratic elution with 37% acetonitrile in water for 31 min, as pre-
viously optimised for these analytes [20] (see Fig. 1). Oligoester
analysis was conducted by elution with 50% ACN held to 5 min, fol-
lowed by a gradient to 100% ACN at 10 min, and 100% ACN held to
15 min.
2. Materials and methods
2.1. Materials
All solvents were of HPLC grade, unless otherwise indicated.
Methanol (MeOH), trifluoroacetic acid (TFA, reagent grade) and
vinyl laurate (GC grade) were from Sigma–Aldrich, acetonitrile
(ACN) was from Panreac Quimica, while N,N-dimethylformamide
(DMF, GC grade) was from Fluka. Water (H2O) of Type I purity (ELGA
PURELAB flex) was degassed by sonication for 20 min (Bransonic
2510E-MT, water bath). Sucrose was from Nordic Sugar A/S, while
molecular sieves (0.3 nm, aluminosilicate beads ∼2 nm) were from
Merck. Enzyme and protein formulationsused were ALP-901 (crude
formulation of ALP-101 from Streptomyces sp., Toyobo), subtilisin
A (Bacillus licheniformis, lyophilised, Novozymes A/S), Alcalase 3.0 T
(granulate of subtilisin A, Novozymes A/S), thermolysin (Bacillus
thermoproteolyticus, lyophilised, Fluka) and casein (technical grade,
Sigma–Aldrich).
Standard curves for sucrose and 6-O-lauroyl sucrose were con-
structed from the analysis of sample series over the respective
ranges of 0.05–60 and 0.06–90 ␮g injected analyte. CAD exhibits
characteristic non-linear mass–response relations expressed by the
equation:
A = aMb
(1)
2.2. Sucrose laurate synthesis reactions
where A is the area response of the detector, M is the injected
mass of analyte, and a and b are values specific to analytes
and chromatographic conditions. Linear regression was applied to
logarithmically transformed peak areas, and from the linearised
mass–response relation:
Sucrose was powdered with mortar and pestle and dried at
120 ◦C for 24 h, while DMF and vinyl laurate were dried over molec-
ular sieves (40 g/L) for at least 24 h. Reactions were performed by
dissolving sucrose (50 mM) in DMF (5 mL) containing molecular
sieves (40 g/L) in serum flasks (25 mL) with stirring magnets for ca.
10 min in a stirring block thermostat (50 ◦C, 600 rpm, VARIOMAG
15.5 with TELEMODUL 40 CT). Enzyme or protein formulation
(10 g/L) was added where appropriate, vinyl laurate (400 mM) was
added and the flasks stoppered with rubber caps and sealed with
parafilm. Reactions with ALP-901, alcalase, and subtilisin were
run in three parallels, and the reaction with thermolysin was
run in two parallels. Control reactions were performed with non-
catalytic protein (casein) and without protein; the reaction with
casein was run in three parallels, as was the reaction without pro-
tein to 6 h, while a single reaction without protein was run to
48 h. The chemical background reaction was accounted for by a
reaction with neither protein nor molecular sieves, performed in
duplicate.
log A = b log M + log a
(2)
the coefficient b and the term log a were determined for each
analyte [21]. By rearranging Eq. (1) and inserting the determined
values, the injected mass of analyte (M, ng) were determined from
peak area (A, pA s) according to the equations:
A1.89
M
2.1% =
(3)
1.565
for sucrose, and
M
3.5% = 4.20A1.09
(4)
for sucrose laurate.
After 2, 4, 6, 24 and 48 h reaction time, samples (100 ␮L)
were withdrawn from the reaction mixture using a syringe (1 mL)
and transferred to centrifuge tubes (1.2 mL). After centrifugation
(14,000 rpm, 3+ min) supernatant aliquots (50 ␮L) were diluted
with methanol (ratio 1:9, 1:2 for the reaction without protein) for
analysis by RP-HPLC.
2.4. Mass spectrometry
Sucrose laurate product (lyophilised, 3.3 mg) in methanol
(100 ␮L) was analysed by matrix-assisted laser desorp-
tion/ionisation time-of-flight (MALDI-TOF) mass spectrometry, as

28
A. Lie et al. / Journal of Molecular Catalysis B: Enzymatic 106 (2014) 26–31
Table 1
Initial rates of formation for 2-O-lauroyl sucrose.
Enzyme formulations (w/mol sieves)
Controls (w/mol sieves)
Control (w/o mol sieves)
Thermolysin
108
ALP-901
Alcalase
Subtilisin
Casein
No protein
3.24 0.63
Background
6
22.2 1.9
1.78 0.17
0.765 0.234
0.981 0.020
0.238 0.019
Initial rates (␮M/min) were calculated from 2-h samples, except for the casein reaction where it was calculated from the 4-h sample.
previously described [19], using DHB (2,5-dihydroxybenzoic acid)
as matrix.
reaction without protein and the concentration of molecular sieves
(40 g/L) corresponded well with their observations. In conjunction
with these results, the lower initial rate with the non-catalytic pro-
tein casein added to the reaction, seemed to be due to the presence
of protein and may be caused by an interaction between protein
and the molecular sieves reducing the catalytic effect.
3. Results and discussion
All protease-catalysed reactions, as well as the control reac-
tions with casein and without protein, resulted in the synthesis
of sucrose laurate. Mass spectrometry of the sucrose laurate prod-
uct gave characteristic signals from the molecular ion adducts with
Na and K (m/z 547 and 563, respectively), corresponding to the
observations of Wang et al. [22] for glucose, fructose and sucrose. 2-
O-Lauroyl sucrose was the most abundant monoester synthesised,
and its concentration increased over the first 6 h of reaction time in
all reactions (see Fig. 2). The initial rates of formation for this regioi-
somer ranged from 0.238 to 108 ␮M/min (see Table 1), with the
control reaction with neither protein nor molecular sieves show-
ing the lowest rate. This reaction was considered as the chemical
background reaction. The three conditions exhibiting highest ini-
tial rates were thermolysin, ALP-901 and no protein, and the rates
with thermolysin and ALP-901 were, respectively, more than 450
and 90 times the rate of the chemical background reaction, and
about 33 and 7 times the rate of the reaction without protein (but
with molecular sieves).
Compared to the remaining reactions where molecular sieves
were included, the initial rate of the reaction without protein was
about 2 times that with alcalase, about 3.5 times the rate with
casein and about 4 times the rate with subtilisin, indicating the
presence of catalytic functionality other than enzyme under these
reaction conditions. Ritthitham et al. [13] showed Celite to have cat-
alytic activity in the reaction between sucrose and vinyl stearate in
DMF/DMSO (1:1, v/v). The initial rate of formation for 2-O-stearoyl
sucrose was reported to be 0.17 ␮M/min in medium without Celite,
1.7 ␮M/min with 10 g/L Celite in the medium, while it increased
to 11.7 ␮M/min with 100 g/L. The molecular sieves employed in
the present investigation were made of sodium aluminosilicate,
a material similar to Celite, and the initial rate observed in the
The lower initial rates of the subtilisin formulations compared
to the reaction without protein, indicated decreased catalytic activ-
ity in these reactions, similar to what was observed in the presence
of the non-catalytic protein, casein. Thus, catalytic activity specifi-
cally ascribable to subtilisin in either of the two formulations could
not be distinguished from that of the non-catalytic protein, casein.
Substrate inhibition, particularly with vinyl laurate, could decrease
the activity of the enzymes, although subtilisin has been reported
to catalyse the acyl-transfer between vinyl laurate and sucrose in
DMF–pyridine (1:1, v/v) at higher acyl donor concentrations than
employed in the present investigation [10]. Subtilisin and other
alkaline proteases have also been reported to exhibit high sta-
bility in pure DMF, in fact, higher than in other commonly used
hydrophilic, aprotic solvents, such as dimethyl sulfoxide (DMSO),
and in some cases higher than in water [13,23], although the gran-
ulate of subtilisin A (Alcalase) was observed to dissolve in the
reaction medium. In addition, the tautomerisation by-product of
acyl transfer with vinyl esters, acetaldehyde, can act as an alkylating
agent through formation of Schiff’s bases with lysine residues in the
protein, which destabilises some enzymes [24], although the pres-
ence of molecular sieves has been indicated as beneficial through
trapping of the acetaldehyde [25].
The concentration of sucrose was observed to decrease progres-
sively during the course of all the reactions, and the total conversion
of sucrose after 48 h ranged from 19 to 96% (see Table 2). The
highest degree of conversion was observed in the reaction with
thermolysin, while the reactions with no protein and ALP-901
showed degrees of conversion above 50%. The sucrose laurate yield
after 48 h ranged from 12 to 53%, and the reaction with ALP-901
resulted in the highest yield. To allow for the complete ester substi-
tution of sucrose, 8 equiv. of acyl donor were used in the reactions.
The yield of sucrose laurate compared to the substrate conversion
in the reactions with thermolysin, ALP-901 and without protein
indicated the formation of di- and oligoesters at the expense of
monoesters. The presence of oligoesters was observed after 48 h in
these reactions, and the detected amounts of mono- and oligoesters
corresponded to the degree of sucrose conversion. Previous reports
suggest that a lower ratio of acyl donor to acyl acceptor would
increase the selectivity towards monoester product [10,26].
25
20
15
10
5
Table 2
Sucrose conversion and sucrose laurate yields after 48 h.
Reaction
Sucrose conversion (%)
Sucrose laurate yield (%)
Thermolysin
ALP-901
Alcalase
Subtilisin
Casein
No protein
96
66
34
19
21
82
3
5
5
8
9
2
29
53
44
12
34
2
4
5
3
9
0
0
1
2
3
4
5
6
reaction time (hours)
Fig. 2. Progress curves for 2-O-lauroyl sucrose. Reaction conditions: (–ꢁ) ther-
molysin, (–᭹) ALP-901, (· · ·ꢀ) alcalase, (–ꢁ) subtilisin, (–♦) casein, (· · ·ꢂ) no protein.
Error bars indicate standard deviation, determined from triplicate reactions, except
for thermolysin (duplicates).
46 10
Conversion expressed as ratio of concentration decrease to initial concentration, and
yield expressed as ratio of detected concentration to theoretical maximum.

A. Lie et al. / Journal of Molecular Catalysis B: Enzymatic 106 (2014) 26–31
29
20
15
10
5
0
Thermolysin
ALP−901
Alcalase
Subtilisin
Casein
No protein
Fig. 3. Catalyst effect on sucrose laurate regioisomer concentration in reaction mixture after 48 h. Regioisomers of sucrose laurate: ( ) 2, ( ) 3, ( ) 4, ( ) 6, ( ) 1ꢀ, ꢀ 3ꢀ.
The 4ꢀ- and 6ꢀ-regioisomers were not detected except in trace amounts in the reaction without protein. Error bars indicate standard deviations, determined from triplicate
reactions, except for thermolysin (duplicates).
a)
b)
50
40
30
20
10
0
25
20
15
10
5
50
40
30
20
10
0
25
20
15
10
5
0
0
0
10
20
30
40
50
0
10
20
30
40
50
reaction time (hours)
reaction time (hours)
Fig. 4. Progress curves for the reactions catalysed by (a) thermolysin and (b) ALP-901. Sucrose: (
); sucrose laurate regioisomers: (–ꢁ) 2, (–᭹) 3, (· · ·ꢀ) 4, (–ꢁ) 6, (–♦) 1ꢀ,
(· · ·ꢂ) 3ꢀ. Error bars indicate standard deviations determined from duplicates for thermolysin and triplicates for ALP-901.
The total concentration of sucrose laurate observed after 48 h
differed between the reaction conditions, however, 2-O-lauroyl
sucrose was the most abundant regioisomer in all reactions (see
Fig. 3). The regioisomeric distribution of sucrose monoesters after
48 h was similar in the reactions with alcalase, subtilisin and casein,
while it differed somewhat in the reactions with ALP-901 and with-
out protein, whose distributions were similar to each other. In the
background control, very low concentrations of only the 2- and
3ꢀ-regioisomer were detected during 6 h of reaction. The reaction
with thermolysin exhibited a unique regioisomeric distribution
with lower proportion of the 3ꢀ-regioisomer and the 4-regioisomer
not detected (see Table 3). The detected appearance of the sucrose
laurate regioisomers largely corresponded to the apparent rates of
formation, based on regioisomeric distribution and concentration
after 48 h.
The reaction progress over 48 h for the two enzyme-catalysed
reactions with the highest initial rates was compared (see Fig. 4).
In the thermolysin-catalysed reaction the highest concentration of
2-O-lauroyl sucrose (23.7 mM) was observed at 24 h (see Fig. 4a).
Other regioisomers detected from 2 h, in descending order of con-
centration, were 3ꢀ, 1ꢀ, 3 and 4 (see Table 3). The 4-regioisomer
concentration peaked at 4 h and could no longer be detected from
24 h. The concentrations of regioisomers 3ꢀ and 1ꢀ peaked at 6 h,
while the concentration of the 3-regioisomer was highest at 24 h.
The overall highest yield of sucrose laurate (63 5%) was detected
in the reaction with thermolysin already after 6 h (see Fig. 4a),
Table 3
Catalyst effect on appearance and distribution of sucrose laurate regioisomers.
Reaction
Reaction time
2 h
Regioisomeric distribution after 48 h (2:3:4:6:1ꢀ:3ꢀ)a
48 h
4 h
6 h
3ꢀ
24 h
Thermolysin
ALP-901
Alcalase
Subtilisin
Casein
No protein
2, 3, 4, 1ꢀ, 3ꢀ
86:8:–:–:4:3
72:5:2:1:7:14
2, 4, 1ꢀ, 3ꢀ
3, 6
2
2
1ꢀ, 3ꢀ
3, 4, 6
4, 1ꢀ
69:4:4:1:8:14
71:3:5:–:7:15
69:4:4:1:7:15
74:5:2:1:7:13
3
6
2, 3ꢀ
3, 4, 1ꢀ
2, 3ꢀ
3, 4, 6, 1ꢀ
4ꢀ, 6ꢀ
b
Regioisomer identity is indicated under the reaction time of first detection for each reaction condition.
a
The regioisomers 4ꢀ and 6ꢀ were only detected in trace amounts in the reaction without protein. Sums of ratios can differ from 100 due to rounding.
The reaction without protein was not sampled after 24 h.
b

30
A. Lie et al. / Journal of Molecular Catalysis B: Enzymatic 106 (2014) 26–31
modelling of molecular electrostatic potential profiles showed the
50
40
30
20
10
0
5
4
3
2
1
0
2-OH substituent to be the most electropositive hydroxyl group,
and experimental confirmation by benzylation in DMF resulted in
the order of reactivity 2-OH > 1ꢀ-OH > 3ꢀ-OH [30]. Acid chloride acy-
lation in water have also confirmed the 2-OH as the most acidic
and reactive substituent [31,32]. Molinier et al. [33] showed that
acyl migration in sucrose monoesters is catalysed by aqueous,
hydrophilic organic media, resulting in rapid conversion from the
2- to the 3-regioisomer, which has also been shown to be catal-
ysed by the protease AL-89 [13], and slower further conversion
to the 6-regioisomer. Even slower migrations favour the primary
hydroxyl groups 6, 1ꢀ and 6ꢀ. The resulting regioisomeric distribu-
tion of sucrose acylation does not necessarily express the relative
reactivity of the hydroxyl groups, but the intricate competition
between esterification and migration [32]. In the current reactions,
the regioisomeric distribution conformed largely to the reactivity
predicted by the electrostatic potential with the major products
being the 2-, 3ꢀ- and 1ꢀ-regioisomers, indicating that acyl migration
was not favoured.
0
2
4
6
reaction time (hours)
Fig. 5. Progress curves for the reaction without protein. Sucrose: (
); sucrose
laurate regioisomers: (–ꢁ) 2, (–᭹) 3, (· · ·ꢀ) 4, (–ꢁ) 6, (–♦) 1ꢀ, (· · ·ꢂ) 3ꢀ. Error bars
indicate standard deviations determined from triplicate reactions.
The distinctly different product compositions after 48 h in the
reaction with thermolysin seemed to be a result of the overall
higher reaction rate and the reaction thus progressing further over
the reaction time studied. The reaction progress over 48 h for ALP-
901 resembled that over 6 h for thermolysin, as supported by the
similarity of the sucrose laurate regioisomeric distribution and
degree of sucrose conversion at these respective reaction times,
and by the decline in the concentration of some regioisomers from
after 4 h with thermolysin and after 24 h with ALP-901.
where the regioisomeric distribution was 71:5:2:9:13 (2:3:4:1ꢀ:3ꢀ,
6, 6ꢀ and 4ꢀ not detected) and the degree of sucrose conversion indi-
cated only monoester product was present. The subsequent decline
in sucrose laurate concentration was due to oligoester formation,
as previously mentioned.
In the reaction catalysed by ALP-901 the concentration of 2-
O-lauroyl sucrose increased progressively during the reaction to
19.0 mM at 48 h (see Fig. 4b). Other regioisomers detected at 2 h,
in descending order of concentration, were 3ꢀ, 1ꢀ and 4, while the
3- and 6-regioisomers were detected from 4 h (see Table 3). The
concentrations of regioisomers 4 and 6 peaked at 24 h, while the
other regioisomers detected were observed in highest concentra-
tion after 48 h of reaction.
The reaction progress over 6 h for the control reaction without
protein was also considered (see Fig. 5). The 2-regioisomer was the
most abundant at 3.00 mM after 6 h, and the regioisomeric distribu-
tion at this time, 69:3:5:1:7:16 (2:3:4:6:1ꢀ:3ꢀ), was very similar to
the distributions observed in the reactions with alcalase, subtilisin
and casein after 48 h (see Table 3). The 3ꢀ-regioisomer was detected
from 2 h and the regioisomers 1ꢀ, 4, 6 and 3, in descending order
of concentration, were detected from 4 h. The concentration of
sucrose varied irregularly between the time points and the standard
deviations were also large, possibly as a result of the samples ana-
lysed containing sucrose concentrations close to the upper limit
of the detector range. After 48 h, all regioisomers were detected in
the sample and the highest concentrations of all regioisomers were
observed, with 17.0 mM of 2-O-lauroyl sucrose detected.
The observed selectivity of both thermolysin and ALP-901 was
predominantly towards formation of the 2-regioisomer. This corre-
sponded to previous reports on thermolysin-catalysed acyl transfer
reactions in hydrophilic, aprotic solvents [19,27], and reports on
catalysis by another alkaline protease, AL-89, in vinyl fatty acid
acyl transfers to sucrose in DMF-DMSO (1:1, v/v) [13]. Previ-
ously reported subtilisin-catalysed reactions with short-chained
acyl donors in various DMF–co-solvent mixtures, have shown
selectivity for acylation at the 1ꢀ-position of sucrose [11,16,28],
however, as outlined above, the catalytic activity of the subtilisin
preparations employed in the present investigation appeared to be
negligible and the regioisomeric distribution. At 48 h the regioiso-
meric distribution of the thermolysin catalysed reaction differed
significantly from the remaining reactions, while the proportion
of the 4-regioisomer of the reactions with ALP-901 and without
protein was half that of the three slower reactions. Nuclear mag-
netic resonance studies of sucrose in DMSO solution have showed
intramolecular hydrogen bonds between 2-OH and both 1ꢀ-OH and
3ꢀ-OH [29], which have conformation-stabilising effects. Computer
4. Conclusion
The acylation of sucrose with vinyl laurate acyl donor in the
hydrophilic, aprotic solvent N,N-dimethylformamide resulted in
the formation of sucrose laurate in yields from 12 to 53% after
48 h under different catalytic conditions. The serine protease ALP-
901 produced the highest yield at this reaction time, while the
reaction with thermolysin achieved the overall highest yield (63%)
after 6 h, with only monoesters synthesised. 2-O-Lauroyl sucrose
was the most abundant monoester regioisomer synthesised, it
was among the first regioisomers to appear and its concentration
increased over the first 6 h in all reactions. The highest concen-
tration of 2-O-lauroyl sucrose observed was 23.7 mM after 24 h in
the thermolysin-catalysed reaction, while in the reaction catalysed
by ALP-901 and in the reaction without protein, it was 19.0 mM
and 17.0 mM after 48 h, respectively. The initial rates of formation
of the reactions catalysed by thermolysin and ALP-901 were more
than 450 and 90 times the rate of the chemical background reaction,
respectively.
The detected appearance of the sucrose laurate regioisomers
largely corresponded to the apparent rates of formation, based on
initial rates of formation for the 2-regioisomer and monoester con-
centration and regioisomeric distribution after 48 h. The observed
sucrose laurate regioisomeric distribution after 48 h (2:3:4:6:1ꢀ:3ꢀ)
was 72:5:2:1:7:14 in the reaction catalysed by ALP-901, and
74:5:2:1:7:13 in the reaction without protein. In the reaction catal-
ysed by thermolysin, the distribution was 71:5:2:–:9:13 after 6 h,
while it was 86:8:–:–:4:3 after 48 h. The concentration of sucrose
laurate was observed to decline with reaction times above 6 h in
the reaction with thermolysin due to oligoester formation.
The reaction between sucrose and vinyl laurate with no protein
in the reaction mixture appeared to be catalysed by aluminosil-
icate molecular sieves present in the reaction medium, with an
initial rate of formation 13 times higher than that of the chemical
background reaction. Non-catalytic protein in the reaction medium
seemed to lower the catalytic activity of the molecular sieves.

A. Lie et al. / Journal of Molecular Catalysis B: Enzymatic 106 (2014) 26–31
31
Acknowledgements
[13] S. Ritthitham, R. Wimmer, A. Stensballe, L.H. Pedersen, J. Mol. Catal. B: Enzyme
59 (2009) 266–273.
[14] P. Degn, W. Zimmermann, Biotechnol. Bioeng. 74 (2001) 483–491.
[15] D. Reyes-Duarte, N. Lopez-Cortes, M. Ferrer, F.J. Plou, A. Ballesteros, Biocatal.
Biotransform. 23 (2005) 19–27.
[16] P. Potier, A. Bouchu, J. Gagnaire, Y. Queneau, Tetrahedron: Asymmetry 12 (2001)
2409–2419.
[17] S. Riva, M. Nonini, G. Ottolina, B. Danieli, Carbohydr. Res. 314 (1998) 259–266.
[18] T. Watanabe, R. Matsue, Y. Honda, M. Kuwahara, Carbohydr. Res. 275 (1995)
215–220.
[19] N.R. Pedersen, P.J. Halling, L.H. Pedersen, R. Wimmer, R. Matthiesen, O.R. Velt-
man, FEBS Lett. 519 (2002) 181–184.
[20] A. Lie, L.H. Pedersen, J. Chromatogr. A 1311 (2013) 127–133.
[21] T. Vehovec, A. Obreza, J. Chromatogr. A 1217 (2010) 1549–1556.
[22] C. Wang, J. Li, S. Yao, Y. Guo, X. Xia, Anal. Chim. Acta 604 (2007) 158–164.
[23] H. Ogino, H. Ishikawa, J. Biosci. Bioeng. 91 (2001) 109–116.
[24] H.K. Weber, H. Stecher, K. Faber, Biotechnol. Lett. 17 (1995) 803–808.
We are grateful to Associate Prof. Allan Stensballe at the Depart-
ment of Health Science and Technology, Aalborg University, for
performing the mass spectrometry. The present research was
funded by the Danish Council for Independent Research.
References
[1] M. Ferrer, J. Soliveri, F.J. Plou, N. Lopez-Cortes, D. Reyes-Duarte, M. Christensen,
J.L. Copa-Patino, A. Ballesteros, Enzyme Microb. Technol. 36 (2005) 391–398.
[2] F.J. Plou, M.A. Cruces, M. Ferrer, G. Fuentes, E. Pastor, M. Bernabe, M. Chris-
tensen, F. Comelles, J.L. Parra, A. Ballesteros, J. Biotechnol. 96 (2002) 55–66.
[3] M. Habulin, S. Sabeder, Z. Knez, J. Supercrit. Fluids 45 (2008) 338–345.
[4] T. Watanabe, S. Katayama, M. Matsubara, Y. Honda, M. Kuwahara, Curr. Micro-
biol. 41 (2000) 210–213.
[5] E. Castillo, F. Pezzotti, A. Navarro, A. Lopez-Munguia, Biotechnol. J. 102 (2003)
251–259.
[6] J.F. Robyt, Essentials of Carbohydrate Chemistry, Springer, New York,
1998.
[7] Y.G. Shi, J.R. Li, Y.H. Chu, J. Chem. Technol. Biotechnol. 86 (2011) 1457–1468.
[8] G. Carrea, S. Riva, Angew. Chem. Int. Ed. 39 (2000) 2226–2254.
[9] N.R. Pedersen, R. Wimmer, J. Emmersen, P. Degn, L.H. Pedersen, Carbohydr. Res.
337 (2002) 1179–1184.
[10] N.R. Pedersen, R. Wimmer, R. Matthiesen, L.H. Pedersen, A. Gessesse, Tetrahe-
dron: Asymmetry 14 (2003) 667–673.
[11] S. Riva, J. Chopineau, A.P.G. Kieboom, A.M. Klibanov, J. Am. Chem. Soc. 110
(1988) 584–589.
[12] O. Almarsson, A.M. Klibanov, Biotechnol. Bioeng. 49 (1996) 87–92.
[25] K. Faber, Biotransformations in Organic Chemistry:
Springer, Heidelberg, 2004.
A Textbook, 5th ed.,
[26] M. Ferrer, M.A. Cruces, M. Bernabe, A. Ballesteros, F.J. Plou, Biotechnol. Bioeng.
65 (1999) 10–16.
[27] I. Perez-Victoria, J.C. Morales, Tetrahedron 62 (2006) 2361–2369.
[28] J.O. Rich, B.A. Bedell, J.S. Dordick, Biotechnol. Bioeng. 45 (1995) 426–434.
[29] J.C. Christofides, D.B. Davies, J. Chem. Soc.: Chem. Commun. (1985) 1533–1534.
[30] F.W. Lichtenthaler, S. Immel, P. Pokinskyj, Liebigs Ann. 1995 (1995) 1939–1947.
[31] Y. Queneau, S. Chambert, C. Besset, R. Cheaib, Carbohydr. Res. 343 (2008)
1999–2009.
[32] S. Thevenet, A. Wernicke, S. Belniak, G. Descotes, A. Bouchu, Y. Queneau, Car-
bohydr. Res. 318 (1999) 52–66.
[33] V. Molinier, K. Wisnienwski, A. Bouchu, J. Fitremann, Y. Queneau, J. Carbohydr.
Chem. 22 (2003) 657–669.