A kinetic study of the lipase-catalyzed ethanolysis of two short-chain triradylglycerols: Alkylglycerols vs. triacylglycerols

Article abstract of DOI:10.1016/j.molcatb.2010.02.010

Lipase-catalyzed ethanolysis of two short-chain triradylglycerols, namely tributyrin and 2,3-dibutyroil-1-O-alkylglycerols, have been studied. Much faster rate of reaction for the ethanolysis of tributyrin than that of 2,3-dibutyroil-1-O-alkylglycerols was attained. A kinetic model for the rate of release of ethyl butyrate and for the inactivation of the lipase has been also studied. The parameter corresponding to the release of ethyl butyrate was one order of magnitude higher for ethanolysis of tributyrin than the corresponding of 2,3-dibutyroil-1-O-alkylglycerols.On the contrary, the stability of Novozym 435 during ethanolysis of 2,3-dibutyroil-1-O-alkylglycerols was higher than the corresponding of tributyrin.At the reaction conditions under study, both ethanolysis reactions take place with high selectivity and yield monoesterified alkylglycerols and sn-2 monobutyrin as the main acylglycerols in the reaction mixtures.

Full text of DOI:10.1016/j.molcatb.2010.02.010

Journal of Molecular Catalysis B: Enzymatic 64 (2010) 101–106
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
A kinetic study of the lipase-catalyzed ethanolysis of two short-chain
triradylglycerols: Alkylglycerols vs. triacylglycerols
Luis Vázqueza, Oscar Fernandeza, Rosa M. Blancob, F. Javier Sen˜oránsa,
Guillermo Regleroa, Carlos F. Torresa,∗
a Sección Departamental Ciencias de la Alimentación, Facultad de Ciencias, Universidad Autónoma de Madrid, 28049 Cantoblanco, Madrid, Spain
b Instituto de Catálisis y Petroleoquímica, CSIC, Marie Curie, 2, 28049 Madrid, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 2 October 2009
Received in revised form 28 January 2010
Accepted 22 February 2010
Available online 10 March 2010
Lipase-catalyzed ethanolysis of two short-chain triradylglycerols, namely tributyrin and 2,3-dibutyroil-
1-O-alkylglycerols, have been studied. Much faster rate of reaction for the ethanolysis of tributyrin than
that of 2,3-dibutyroil-1-O-alkylglycerols was attained. A kinetic model for the rate of release of ethyl
butyrate and for the inactivation of the lipase has been also studied. The parameter corresponding to
the release of ethyl butyrate was one order of magnitude higher for ethanolysis of tributyrin than the
corresponding of 2,3-dibutyroil-1-O-alkylglycerols.
On the contrary, the stability of Novozym 435 during ethanolysis of 2,3-dibutyroil-1-O-alkylglycerols
was higher than the corresponding of tributyrin.
At the reaction conditions under study, both ethanolysis reactions take place with high selectivity
and yield monoesterified alkylglycerols and sn-2 monobutyrin as the main acylglycerols in the reaction
mixtures.
Keywords:
Candida antarctica
Alkylglycerols
Ethanolysis
Butyric
Lipase
Kinetics model
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
be due to its synergistic action with another antiproliferative agent,
1,25-dihydroxyvitamin D3 [dihydroxycholecalciferol; (OH)2D3]. In
Alkylglycerols, alkylglycerophospholipids and their derivatives,
commonly known as ether lipids, have been the subject of much
attentionbecause of theirspecial physiologicalfunctions in humans
[1,2]. Ether lipids have been used in the therapy of cancer [3], since
they are potent antineoplastic agents which inhibit growth, show
antimetastatic activity and induce differentiation and apoptosis in
cancer cells [4,5]. Immune stimulators properties have also been
attributed to dietary ingestion of these substances [6].
In recent years, numerous synthetic ether lipids that do not
occur in nature have been prepared to find therapeutic agents
such as agonists and antagonists of natural lipid mediators [7]. One
example of the mentioned synthetic ether lipids is PAF mimics that
retain a short-chain residue esterified at the sn-2 position [8].
The studies done during the last decade provide multiple lines
of evidence that butyrate indeed interferes with the pathogenesis
of colorectal cancer. Butyrate inhibits DNA synthesis and arrests
growth of neoplastic colonocytes [9], modifies expression of genes
involved in chemotherapy resistance [10] and in cell prolifera-
tion/differentiation [11,12], and induces apoptosis [13]. At least
some of butyrate’s antineoplastic effects in colon cancer cells may
various cancer cell lines it has been shown that butyrate and
(OH)2D3 act synergistically in reducing proliferation and enhancing
differentiation of neoplastic cells [14,15].
In spite of its early promise, butyrate is not among the drugs
used for cancer treatment. The major problem has been to achieve
and maintain its millimolar concentrations in blood. Butyrate is
metabolized rapidly as soon as it enters the colonocyte via its active
transport system [16–18], and its plasma concentrations are far
below those required to exert its antiproliferative/differentiating
actions. Prodrug of natural butyrate, such as tributyrin (TB) and
2,3-dibutyroil-1-O-alkylglycerols (SCAKG), are neutral short-chain
fatty acid triradylglycerols that are likely to overcome the pharma-
cokinetic drawbacks of natural butyrate as a drug [19].
In addition, recent studies indicate that butyrate and fish oil
work coordinately to protect against colon tumorigenesis [20,21]
Based on these findings, designing new lipidic vehicles structures
in which these two fatty acids are incorporated into the body
could influence the location of their release and it may allow
targeting of these bioactive molecules to specific areas of other
tissues.
One of the problems that arise is to elucidate which of these
two lipidic vehicles, namely alkylglycerols and triacylglycerols, is
more efficient and chemically stable in plasma, diffuses through
biological membranes and is metabolized by intracellular lipases,
∗
Corresponding author. Tel.: +34 91 4973091; fax: +34 91 4978255.
E-mail address: carlos.torres@uam.es (C.F. Torres).
1381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.02.010

102
L. Vázquez et al. / Journal of Molecular Catalysis B: Enzymatic 64 (2010) 101–106
releasing therapeutically effective fatty acids over time directly into
the cell.
Separations were performed on a Hewlett-Packard 5890 series
II gas chromatograph with on-column injection using a 7-m 5%
phenyl methyl silicone capillary column (Quadrex Corporation,
New Haven, CT) (0.25 ␮m i.d.). 12 cm of a 530 ␮m i.d. deactivated
column was used as pre-column. Injector and detector temperature
was 43 ◦C and 360 ◦C respectively. The temperature program was as
follows: starting at 40 ◦C and then heating to 250 ◦C at 42 ◦C min−1
with 10 min hold, followed by heating from 250 ◦C to 325 ◦C at
7.5 ◦C min−1 with 30 min hold. Helium was used as a carrier gas at a
pressure of 5.2 psi. The peaks were computed using GC chemstation
software (Agilent Technologies, Santa Clara, CA).
In a first attempt, the present study focuses on the kinetics of the
lipase-catalyzed ethanolysis of these two lipidic vehicles contain-
ing butyric acid residues. The results obtained fulfill two objectives:
(1) show differences in rate of the reaction, selectivity, and provide
mechanistic information regarding how these two compounds are
recognized by lipases, and (2) lipase-catalyzed ethanolysis of these
two triradylglycerol is very useful step for selectively remove fatty
acid residues at sn-1, and sn-3 positions and it can be used for sub-
sequent production of structured lipids containing short chain and
polyunsaturated fatty acids in a single molecule.
2.5. Mathematical modeling
2. Experimental
2.5.1. Reaction rate for the ethanolysis reaction: uniresponse
model
2.1. Materials
Rate expressions based on a generalized Michaelis–Menten
mechanism for the ethanolysis reaction were utilized (Fig. 1). A
ping–pong mechanism controlled by the rate of deacylation of
the enzyme was proposed. These rate expressions are similar to
those described by Malcata et al. [24] and by Lessard and Hill [25]
for lipase-catalyzed hydrolysis reactions. In particular a modifica-
tion of these rate expressions, presented by Torres et al. [26], was
employed. The main differences with respect to the model pro-
posed by Malcata et al. are: (1) the rate expressions contain the
concentration of ethanol rather than the concentration of water;
(2) a term for the reverse of the ethanolysis reaction is incorpo-
rated in the rate expressions. However, in the present study the
reverse of the ethanolysis reaction can be neglected based on the
findings that nearly 100% of fatty acid ethyl esters can be achieved
in similar systems [27,28]. Hence, the rate expression of the model
is represented by
2,3-Dibutyroil-1-O-octadecylglycerol (SCAKG) was synthesized
according to a methodology previously described by our group
[22]. Tributyrin (TB) was purchased from Sigma–Aldrich (St. Louis,
MO, USA). N-hexadecane for synthesis was purchased from Merck
(Darmstadt, Germany). All solvents used were HPLC grade from
Lab-Scan (Dublin, Ireland). The immobilized lipase B from Candida
antarctica (Novozym 435) was a gift from Novozymes Spain. Meso-
porous silica MS3030 was a kind gift from Silica PQ Corporation
(PA, USA).
2.2. Ethanolysis reaction
TB or SCAKG (1.7 g), ethanol (5.3 g) and n-hexadecane as inter-
nal standard (0.2 g) were added to a 100-mL flask and mixed by
swirling. Then, the lipase 10% by weight was added. The flask was
placed in an orbital shaker at 50 ◦C and 200 rpm. Samples of 50 ␮L
were withdrawn periodically. The reaction was allowed to proceed
for 50 min to 420 min.
d[G]
(vmax/Km)[G][B]
=
(1)
dt
1 + ([G]/Km) + ([P]/Ki)
[G] and [P] are the molar concentrations of the parent acylglycerol,
and the lower acylglycerol, respectively. [B] denotes the concen-
tration of ethanol, the vmax is the maximum rate of reaction at a
particular glyceride bond [G] at a saturating concentration of the
substrate (and in the absence of the other glycerides), the Km is
the Michaelis–Menten constant for the glycerides and the Ki is the
inhibition constant for the corresponding ethyl esters. The lumped
parameters, vmax, Ki, and Km are defined as
2.3. Immobilization of Candida antarctica lipase
The protocol of immobilization of lipase has been previously
described [23]. Silica MS 3030 was functionalized by reaction with
octyltriethoxysilane in toluene, and the octyl silica obtained was
used as the carrier.
1 mL of ethanol was added for each 100 mg octyl silica and the
mixture was left to equilibrate in a closed vial for at least 10 min.
5 mL of the commercial extract of C. antarctica lipase B contain-
ing 20 mg protein were dissolved in 25 mM phosphate buffer pH
7.0, up to a total volume of 10 mL. After assaying the esterasic
activity (hydrolysis of p-nitrophenyl acetate followed spectropho-
tometrically at 348 nm), the enzyme solution was added on the
vial containing the silica in ethanol and kept in mild stirring with
a helical stirrer. Aliquots from suspension and supernatant were
withdrawn at 15 min intervals to assay their respective esterasic
activities. Final time was determined by the lack of activity, or low
constant activity of the supernatant. The suspensions were then
filtered and washed three times with 10 mL volumes of 200 mM
phosphate buffer. The supernatants of the immobilization mix-
tures, as well as the liquids from these washings, were tested for
protein desorption by assays of their catalytic activities (esterasic
activity). After the last washing, the derivatives were twice sus-
pended in 10 mL dry acetone, filtered out and finally vacuum dried
for at least 30 min to ensure a complete drying of the catalyst.
k kp[ET ]
k2p + k
vmax
=
(2)
(3)
(4)
−p
k3
Ki =
k
−3
k
k
k (k−p1+−kp
)
Km
=
−p
1
Fig. 1. Schematic representation of the ethanolysis reaction. E = enzyme, G = glyceryl
ester bond (reactant glyceride), E–G = activated form of the enzyme and the native
ester bond, F–G = covalent transformed form of the enzyme and the native ester
bond, E–Q = activated form of the enzyme and the fatty acid ethyl ester, B = ethanol,
2.4. Analyses of the reaction products by gas chromatography
The samples were diluted with chloroform to obtain a final con-
centration lower than 0.1 mg/mL for the subsequent analysis by gas
chromatography.
P = lower glyceride liberated from the native ester, Q = fatty acid ethyl ester. k1, k
k2, k−2, kp, k−p, k3, k−3, are rate constants.
,
−1

L. Vázquez et al. / Journal of Molecular Catalysis B: Enzymatic 64 (2010) 101–106
103
Table 1
Definitions of model parameters in terms of the intrinsic rate constants.
reaction time for a particular batch is defined as
ꢂ
t
ˆ
ˆ
a{t}dt
Parameter
Mechanistic interpretation
t
1
t∗
=
(8)
k
k
k
[E ]/k
ꢀ
k
a0
p
1
2
T
−1 −2
˝
2
ꢀ
ꢁ
k
k
k
k
k
K1
+
−
where t* is the pseudo reaction time, t is a dummy variable, t1 is
the time at which processing of the particular batch of interest was
initiated, and (t − t1) is the (elapsed) time during which a particular
batch of oil is in contact with the lipase (the actual reaction time).
In physical terms when the rate of deactivation of the enzyme is
negligible, the pseudo reaction time t* becomes equal to the actual
reaction time (t − t1). When the rate of deactivation is significant,
the difference between the actual reaction time and the pseudo
reaction time indicates how much longer the reaction mixture must
remain in the reactor to achieve the conversion that would have
been achieved if the enzyme had not been partially deactivated. In
terms of Eqs. (7) and (8):
1
ꢀ
−3
1
1
2
ˆ
k
k
k
3
−1
−1 −2
k
ꢀ
1 + −3 [G0]
k
3
Eq. (1) can be simplified by reparameterization following the
approach of Malcata et al. [24] to yield:
d[G]
dt
˝2[G][B]
1 + K1[G]
−
=
(5)
The corresponding lumped parameters (˝2 and K1) are summa-
rized in Table 1.
ꢃꢄ
ꢃ
ꢅꢆ
ꢄ
ꢃ
ꢅꢆꢅ
1
kd
t∗
=
exp −kd(t1 − t0)
−
exp −kd(t − t0)
(9)
2.5.2. Loss of enzyme activity with time
Knowledge of the rate law governing the deactivation process is
often important in modeling enzyme-catalyzed processes. For the
purpose of modeling, deactivation is usually treated as if it occurs
in a single step. It is often assumed that irreversible deactivation
of the active enzyme (E) to an inactive form (Ed) obeys first order
kinetics:
Note that as kd approaches zero, the difference in the exponen-
tial terms approaches kd(t − t1) and t* approaches (t − t1). The
reader should also note that the value of (t1 − t0) represents the
sum of the times that the enzyme was used as a biocatalyst in
previous trials and the total time necessary to recover the cata-
lyst in the various filtration steps prior to its use in the trial of
interest.
k
d
E−→ Ed
(6)
The manner in which the enzyme activity varies with time is then
given by
3. Results and discussion
3.1. Direct comparison between lipase-catalyzed ethanolysis of
tributyrin and sn-2,3-dibutyroil-1-O-alkylglycerols
a{t} = a exp −k (t − t )
(7)
[
]
0
0
d
where t is time; t0 is the time at which the first experiment with
a particular sample of enzyme is started; a{t} is the activity of the
enzyme at a particular time t; a0 is the activity of the enzyme at time
t0; and kd is the first order deactivation rate constant. The param-
eter a0 can arbitrarily be assigned to a value of 1 and subsequent
activities are thus normalized with respect to this initial activity.
One of the problems that arise when one tries to evaluate the
loss of enzyme activity in batch reactors is the fact that the time
of the reaction and the time at which the set of experiments was
started are equivalent. This fact does not permit one to separate
the rate of the reaction from the rate at which the enzyme is losing
activity. In order to overcome this problem, data sets were obtained
using two cycles of reaction with the same batch of lipase in order
to distinguish between the times characteristic of the two types
of reaction (ethanolysis and deactivation). Each cycle involves the
same elapsed time for the ethanolysis reaction but for one step in
the cycle, the elapsed time will differ from the time during which
the lipase has been susceptible to deactivation (t − t0). The math-
ematical model used in the analysis of the data thus can account
for the effect of deactivation of the enzyme in the combined rate
expressions for the rates of ethanolysis reaction.
The weight percentage of ethyl butyrate from the reaction mix-
tures released during the first 40 min of both ethanolysis reactions
is shown in Fig. 2. It can be observed that the rate of lipase-catalyzed
ethanolysis of TB is much faster than that of SCAKG. A higher reac-
tion rate can be expected in the case of tributyrin compared to that
of SCAKG because of the fact that TB contains three butyric acid
residues and SCAKG only two. However, inspection of the course
of the reaction indicates that after 15 min of reaction, the percent-
age of ethyl butyrate released is ca. 10% and 50% for ethanolysis
of SCAKG and TB respectively. These results indicate that lipase-
To characterize the loss of enzyme activity, the variable of
importance is the time elapsed since the start of the experiments
(t − t0). The amounts of the butyric acid ethyl ester obtained depend
not only on the time the reactants in a particular batch are in con-
tact with the lipase, but also the enzymatic activity profile during
the time that batch is being processed. Graphical depictions of the
resulting functional dependence thus require the use of a three
dimensional plot. However, in order to obtain plots that can be
more readily interpreted, it is convenient to employ the concept
of a pseudo reaction time in order to permit representation of the
data in two dimensions [24]. The pseudo reaction time represents
a mathematical combination of the time elapsed since the start of
the series of experiments and the real reaction time. The pseudo
Fig. 2. Comparison between the rate of release of ethyl butyrate in lipase-catalyzed
ethanolysis of tributyrin and short-chain alkylglycerols. Filled symbols correspond
to tributyrin, open symbols correspond to short-chain alkylglycerols.

104
L. Vázquez et al. / Journal of Molecular Catalysis B: Enzymatic 64 (2010) 101–106
vation of the lipase and were utilized to determine the inactivation
constant according to the methodology described in Section 2.5.
Besides the differences in reaction rates between ethanolysis of
SCAKG and TB discussed before, it can be observed that monoes-
terified alkylglycerols reach a plateau after 60 min of ethanolysis
reaction that it is maintained almost constant during the course of
the reaction. This result indicates that at these reaction conditions,
Novozym 435 is highly selective for butyric acid residues located
at sn-3 position. On the contrary, this plateau is not observed in
the case of ethanolysis of TB. In this case, 1,3-dibutyrin is rapidly
transformed in 2-monobutyrin. This observation is consistent with
that by Piyatheerawong et al. [29].
Inspection of Fig. 3A and B again indicates the difference in reac-
tion rates between the two ethanolysis reactions studied. Hence,
the percentage of SCAKG and TB after 20 min of reaction was ca.
50% and below 1%, respectively.
3.3. Kinetics modeling of the lipase-catalyzed ethanolysis of
SCAKG and TB
It should be noted that the kinetic model utilized in the present
study only includes parameters for the forward reaction and inhi-
bition effects. Hence, to avoid overparameterization, facilitate the
non-linear regression analysis and based on the findings that nearly
100% of fatty acid ethyl esters can be achieved in similar ethanol-
ysis reaction systems, Eq. (5) does not account for reversibility of
the reaction that was considered to be negligible.
In this study, in an attempt to improve the results of the ethanol-
ysis reaction towards SCAKG, a second lipase (a lipase from C.
antarctica immobilized on mesoporous silica) was also utilized for
the ethanolysis of SCAKG. The kinetic parameters obtained are
shown in Table 2.
Inspection of the Table 2 indicates that the parameter cor-
responding to release of butyric ethyl ester (˝2) is an order of
magnitude smaller for ethanolysis of SCAKG than the correspond-
ing for ethanolysis of TB. A higher value of this parameter was
attained with the immobilized C. antarctica on mesoporous silica
which indicates that this lipase is slightly faster than Novozym 435
in ethanolysis of SCAKG.
Negative values for K1 were obtained in the three cases under
study. Because the lumped parameter K1 involves a difference of
two terms, namely, 1/Km and 1/Ki (Eq. (1)), negative values can be
obtained for K1.
Fig. 3. Lipase-catalyzed ethanolysis of short-chain alkylglycerols (A) and tributyrin
(B). (A) Open triangles: SCAKG; open circles: monoesterified alkylglycerols, filled
circles: ethyl butyrate, filled triangles: non-esterified alkylglycerols. (B) Open trian-
gles: tributyrin, open circles: dibutyrin, filled circles: ethyl butyrate, filled triangles:
monobutyrin.
Regarding the inactivation of lipases the lowest value of kd was
obtained for ethanolysis of SCAKG catalyzed by Novozym 435. In
this case, the contribution to the fitting of kd was almost negligible
and similar results were attained removing this parameter from the
model. This result indicates that C. antarctica lipase is very stable
at this reaction conditions, and numerous trials reusing the same
batch of lipase should be necessary to get enough data to charac-
terize the activity decay and properly calculate the value of kd for
this lipase. For that reason the value reported that corresponds only
to two consecutive trials, it is an estimate that should be used for
comparison purposes but not to determine the extent of the lipase
inactivation. Unfortunately, a higher value of kd was observed for
ethanolysis of SCAKG catalyzed by the immobilized C. antarctica on
mesoporous silica.
catalyzed ethanolysis of TB is approximately five times faster than
lipase-catalyzed ethanolysis of SCAKG.
3.2. Time course of the ethanolysis of SCAKG and TB in the
presence of Novozym 435
The course of the ethanolysis of SCAKG and TB in the presence
of Novozym 435 is depicted in Fig. 3A and B, respectively. Two tri-
als reutilizing the same batch of lipase were effected in both cases.
The differences observed between the two trials assayed for each
ethanolysis reaction could be, in part, attributed to partial inacti-
Table 2
Parameter estimates for the best fits of lipase-catalyzed ethanolysis of two short-chain alkylglycerols and triacylglycerols.
Parameter
SCAKG
SCAKG
TB
Novozym 435
Immobilized C. antarctica
Novozym 435
˝
K1
kd
4.73E−06 1.94E−06
−1.38E−03 9.05E−08
3.82E−03 1.71E−03
1.17E−05 1.16E−06
−1.34E−03 5.08E−05
2.22E−02 2.66E−03
3.29E−05 7.22E−06
−4.93E−04 1.62E−08
7.60E−02 8.08E−03
2

L. Vázquez et al. / Journal of Molecular Catalysis B: Enzymatic 64 (2010) 101–106
105
These results indicate that the immobilization process on C.
antarctica lipase has two opposite effects: (1) increases the rate of
release of butyric acid ethyl ester (higher value of ˝2) and (2) has
a deleterious effect on the stability of the biocatalyst (lower value
of kd in comparison with that attained with Novozym 435).
In addition, inspection of Table 1 also indicates a relatively high
deactivation constant for ethanolysis of TB that can be attributed to
an additional deleterious effect caused by tributyrin that does not
take place in ethanolysis of SCAKG.
The estimated half-life of the lipases under study according to
the values of kd attained are 180, 35, and 10 h for ethanolysis of
SCAKG catalyzed by Novozym 435, ethanolysis of SCAKG catalyzed
by the immobilized C. antarctica on mesoporous silica, and ethanol-
ysis of TB catalyzed by Novozym 435 respectively.
3.4. Kinetic model of the release of ethyl butyrate
Curves corresponding to fits of kinetics model for release of ethyl
butyrate are shown in Fig. 4.
The model contains the assumption that deactivation of the
enzyme (E) obeys first order kinetics. Hence, the kinetics model
corresponds to combination of Eq. (5) with a first order deac-
tivation term. The reader should also note that the data points
correspond to two different trials in which the same enzyme was
reused (after recovery from the previous trial). To facilitate inspec-
tion of the data, different symbols have been utilized for the data
from the two different trials considered in the analysis. The data
points corresponding to the same elapsed time within a given
trial are characterized by a shift on the pseudo reaction time scale
that depends on the extent of inactivation of the lipase. It can be
observed that no shift in the pseudo reaction time was observed
for the shorter reaction times. The extent of inactivation can only
be noticeable at the longer reaction times. Hence, for example for
the ethanolysis of SCAKG catalyzed by C. antarctica lipase immo-
bilized in mesoporous silica (Fig. 4B), the reaction time 420 min is
shifted to a pseudo reaction time of 392 and 340 min for the first
and the second trial respectively. On the contrary, for the ethanol-
ysis of SCAKG catalyzed by Novozym 435 (Fig. 4A) no significant
shift in the pseudo reaction time is observed at 30 min of reaction
time for both trials. On the contrary even higher lipase activity is
observed at short reaction times which could indicates that other
factors (such as recovery of the lipase and experimental errors)
could partially interfere with the reported value of kd. It should
be also indicated that the extent of the inactivation, to be percep-
tible, should be significantly higher than the experimental error in
each trial.
For these reasons (very short reaction times and experimen-
tal errors) it is not possible to graphically visualize the extent of
inactivation for the ethanolysis of TB catalyzed by Novozym 435.
Inspection of Fig. 4A indicates that ethanolysis of SCAKG cat-
alyzed by Novozym 435 occurs in two stages: a first rapid release
of ethyl butyrate during the first hour of reaction followed by a sec-
ond stage with slower release of ethyl butyrate. This effect could be
attributed to inhibition effect for accumulation of monoesterified
alkylglycerols or ethyl butyrate in the surroundings of the active
site of Novozym 435. This phenomenon was not clearly observed
neither for ethanolysis of SCAKG catalyzed by C. antarctica lipase
immobilized in mesoporous silica (Fig. 4B) nor for ethanolysis of
TB catalyzed by Novozym 435 (Fig. 4C).
Fig. 4. Curves corresponding to fit of kinetics model for lipase-catalyzed ethanolysis
of short-chain alkylglycerols in the presence of Novozym 435 (A), in the presence of
Candida antarctica immobilized on mesoporous silica (B) and ethanolysis of tribu-
tyrin in the presence of Novozym 435 (C). Dashed lines correspond to fitted values.
Open symbols correspond to trial 1 and filled symbols correspond to trial 2.
4. Conclusions
A very efficient methodology for the ethanolysis of two different
triradylglycerols has been attained. Much faster rate of ethanolysis
for TB than for SCAKG was observed.

106
L. Vázquez et al. / Journal of Molecular Catalysis B: Enzymatic 64 (2010) 101–106
High regio-selectivity of Novozym 435 was observed for the
[2] P.T. Pugliese, K. Jordan, H. Cederberg, J. Brohult, J. Altern. Complement Med. 4
(1998) 87–99.
ethanolysis of SCAKG with a rapid accumulation of monoesterified
alkylglycerols in the reaction mixture that is maintained almost
constant the remaining time of the reaction.
The kinetic model developed in this study indicates that the
parameter corresponding to release of butyric ethyl ester (˝2) is
oneorder ofmagnitudehigherfor theethanolysis ofTB compared to
that for the ethanolysis of SCAKG which confirms the experimental
observation of the time course of both reactions.
Finally, low inactivation rate was observed for ethanolysis of
SCAKG in the presence of Novozym 435 (kd = 3.82E−03). Based on
this parameter, the estimated half-life of the lipase is 180 h, which
permit to reutilize the same batch of lipase up to 20 trials.
In conclusion the kinetic model developed in the present study
indicates differences in rate of the reaction, selectivity, and provide
mechanistic information regarding how these two triradylglycerols
are recognized by lipases. In addition, lipase-catalyzed ethanolysis
of these two compounds is very useful step for selectively remove
fatty acid residues at sn-1, and sn-3 positions and it can be used for
subsequent production of structured lipids containing short chain
and polyunsaturated fatty acids in a single molecule.
[3] R. Andreesen, Prog. Biochem. Pharmacol. 22 (1988) 118–131.
[4] W.E. Berdel, Br. J. Cancer 64 (1991) 208–211.
[5] L. Diomede, F. Colotta, B. Piovani, F. Re, E.J. Modest, M. Salmona, Int. J. Cancer
53 (1993) 124–130.
[6] J. Palmblad, J. Samuelsson, J. Brohult, Scand. J. Clin. Lab. Invest. 50 (1990)
363–370.
[7] K. Hartvigsen, A. Ravandi, R. Harkewicz, H. Kamido, K. Bukhave, G. Hølmer, A.
Kuksis, Lipids 41 (2006) 679–693.
[8] A. Tokumura, Prog. Lipid Res. 34 (1995) 151–184.
[9] A. Toscani, D.R. Soprano, K.J. Soprano, Oncogene Res. (1988) 223–238.
[10] L.A. Mickley, S.E. Bates, N.D. Richert, S. Currier, S. Tanaka, F. Foss, N. Rosen, A.T.
Fojo, J. Biol. Chem. 264 (1989) 18031–18040.
[11] C.W. Taylor, Y.S. Kim, K.E. Childress-Fields, L.C. Yeoman, Cancer Lett. 62 (1992)
95–105.
[12] A. Wachtershauser, S.M. Loitsch, J. Stein, Biochem. Biophys. Res. Commun. 272
(2000) 380–385.
[13] A. Hague, G.D. Diaz, D.J. Hicks, S. Krajewski, J.C. Reed, C. Paraskeva, Int. J. Cancer
72 (1997) 898–905.
[14] Y. Tanaka, K.K. Bush, T. Eguchi, N. Ikekawa, T. Taguchi, Y. Kobayashi, P.J. Higgins,
Arch. Biochem. Biophys. 276 (1990) 415–423.
[15] M. Yoshida, Y. Tanaka, T. Eguchi, N.I. kekawa, N. Saijo, Anticancer Res. 12 (1992)
1947–1952.
[16] J. Stein, O. Schröder, V. Milovic, W.F. Caspary, Gastroenterology 108 (1995)
673–679.
[17] O. Schröder, J. Opritz, J. Stein, Digestion 62 (2000) 152–158.
[18] J. Stein, M. Zores, O. Schröder, Eur. J. Nutr. 39 (2000) 121–125.
[19] C. Schröder, K. Eckert, H.R. Maurer, Int. J. Oncol. 13 (1998) 1335–1340.
[20] N. Satya Sree Kolar, R. Barhoumi, J.R. Lupton, R.S. Chapkin, Cancer Res. 67 (2007)
5561–5568.
Acknowledgements
[21] N. Yeevoon, R. Barhoumi, R.B. Tjalkens, Y. Fan, S. Kolar, N. Wang, J.R. Lupton,
R.S. Chapkin, Carcinogenesis 26 (2005) 1914–1921.
[22] C.F. Torres, L. Vazquez, F.J. Sen˜orans, G. Reglero, Process Biochem. 44 (2009)
1025–1031.
This work was supported by the projects AGL2006-02031/ALI
and AGL2008-05655 by Ministerio de Ciencia (Spain) and also
by Comunidad Autonoma de Madrid (ALIBIRD, project number
S-505/AGR-0153) and Consolider-Ingenio FUN-C-FOOD (CSD2007-
00063). A predoctoral fellowship for Luis Vázquez, provided by
Comunidad Autonoma de Madrid, is also acknowledged.
[23] R.M. Blanco, P. Terreros, N. Mun˜oz, E. Serra, J. Mol. Catal. B: Enzym. 47 (2007)
13–20.
[24] F.X. Malcata, C.G. Hill Jr., C.H. Amundson, Biotechnol. Bioeng. 39 (1992)
984–1001.
[25] L.P. Lessard, C.G. Hill Jr., Biotechnol. Bioeng. 69 (2000) 183–195.
[26] C.F. Torres, M. Moeljadi, C.G. Hill Jr., Biotechnol. Bioeng. 83 (2003) 274–281.
[27] Y. Yesiloglu, J. Am. Oil Chem. Soc. 81 (2004) 157–160.
[28] Y. Shimada, Y. Watanabe, A. Sugihara, Y. Tominaga, J. Mol. Catal. B: Enzym. 17
(2002) 133–142.
References
[1] H.K. Mangold, in: F.L. Snyder (Ed.), Ether Lipids. Chemistry and biology, Aca-
demic, New York, 1972, pp. 157–176.
[29] W. Piyatheerawong, Y. Iwasaki, T. Yamane, J. Chromatogr. A 1068 (2005)
243–248.