Evaluation of Rhizopus oryzae lipase for the determination of regiodistribution in triacylglycerols with medium chain fatty acids

Article abstract of DOI:10.1007/s11746-011-1884-y

The nutritional profile and rheological behaviors of lipids is both due to fatty acid composition and regiodistribution on external and internal positions of triacylglycerol. Actual methods for regiodistribution analysis having some restrictions, there is still a need for investigating a safe, simple and environmentally friendly method for the sn-2 position analysis that could especially be used for the analysis of fats containing medium and short chain fatty acids. The objective of this study was to evaluate the 1,3-selectivity and typoselectivity of Rhizopus oryzae lipase in the presence of short/medium chain fatty acids in partial hydrolysis conditions used for regiodistribution analysis. Structured triacylglycerols containing eight-carbon-chain length fatty acids in the sn -2 position were chemically synthesized using DCC/DMAP coupling agent and purification steps by flash-chromatography. The final product showed very high purity and was used as the substrate for 1,3-selectivity evaluation. Typoselectivity was assessed by investigating partial hydrolysis of equimolar blends of homogeneous TAG. This study confirmed the 1,3-selectivity of Rhizopus oryzae lipase in the hydrolysis conditions used, and revealed that this lipase was less influenced by fatty acids chain length than pancreatic lipase. Considering this, Rhizopus oryzae lipase appeared to be a good candidate for regiodistribution analysis of fats containing medium and short chain fatty acids.

Full text of DOI:10.1007/s11746-011-1884-y

J Am Oil Chem Soc (2012) 89:89–96
DOI 10.1007/s11746-011-1884-y
ORIGINAL PAPER
Evaluation of Rhizopus oryzae Lipase for the Determination
of Regiodistribution in Triacylglycerols with Medium Chain
Fatty Acids
•
Marlene Perignon Jerome Lecomte Michel Pina
•
•
`
Anne Renault Camille Simonneau-Deve
´
´ ˆ
•
•
Pierre Villeneuve
Received: 17 December 2010 / Revised: 4 May 2011 / Accepted: 9 June 2011 / Published online: 24 June 2011
Ó AOCS 2011
Abstract The nutritional profile and rheological behav-
iors of lipids is both due to fatty acid composition and
regiodistribution on external and internal positions of tri-
acylglycerol. Actual methods for regiodistribution analysis
having some restrictions, there is still a need for investi-
gating a safe, simple and environmentally friendly method
for the sn-2 position analysis that could especially be used
for the analysis of fats containing medium and short chain
fatty acids. The objective of this study was to evaluate
the 1,3-selectivity and typoselectivity of Rhizopus oryzae
lipase in the presence of short/medium chain fatty acids in
partial hydrolysis conditions used for regiodistribution
analysis. Structured triacylglycerols containing eight-
carbon-chain length fatty acids in the sn-2 position were
chemically synthesized using DCC/DMAP coupling agent
and purification steps by flash-chromatography. The final
product showed very high purity and was used as the
substrate for 1,3-selectivity evaluation. Typoselectivity was
assessed by investigating partial hydrolysis of equimolar
blends of homogeneous TAG. This study confirmed the
1,3-selectivity of Rhizopus oryzae lipase in the hydrolysis
conditions used, and revealed that this lipase was less
influenced by fatty acids chain length than pancreatic
lipase. Considering this, Rhizopus oryzae lipase appeared
to be a good candidate for regiodistribution analysis of fats
containing medium and short chain fatty acids.
Keywords Structured triacylglycerols ꢀ
Regiodistribution ꢀ Rhizopus oryzae lipase ꢀ
Typoselectivity
Introduction
There is currently a significant interest in chemical or
enzymatic processes to modify the composition and/or
regiodistribution of oils and fats to obtain new lipids, known
as structured triacylglycerols (structured TAG) [1]. Those
trends are mainly due to a growing concern for nutritional
impact and FA bioavailability which are in fact governed
both by the overall fatty acid (FA) composition and the
stereospecific distribution of FA in TAG molecules [2].
This determination of FA regiodistribution on TAG requires
methods that can be accurately employed on a wide range of
oils and fats substrates which may greatly differ in term of
fatty acid composition. Many various methods have then
been developed such as chemical procedures using the
action of Grignard reagents that react in a non-specific way
on ester functions of triacylglycerols. Thus, Brockerhoff
et al. [3] analyze the ab-diacylglycerols (ab-DAG) obtained
after partial deacylation by Grignard reagent. However, the
subsequent calculation of internal position uses factors
which overemphasize the experimental error [4]. Therefore,
Turon et al. [5] developed a modification of the Brockerhoff
procedure in which the analysis is carried out on
a-monoacylglycerols which are easier to collect than dia-
cylglycerols isomers after subsequent TLC elution. How-
ever, methods involving Grignard reagents suffer various
drawbacks. Indeed, these methods use an ethyl magnesium
bromide which is a toxic chemical, highly reactive and very
sensitive to water. Furthermore, those chemical methods are
tedious and their accuracy is reliant on the limitation of the
´
M. Perignon ꢀ J. Lecomte ꢀ M. Pina ꢀ P. Villeneuve (&)
´
`
CIRAD, UMR Ingenierie des Agro-polymeres et Technologies
Emergentes, Campus Supagro/INRA, 2 place Viala,
34060 Montpellier, France
e-mail: pierre.villeneuve@cirad.fr
A. Renault ꢀ C. Simonneau-Deve
St Hubert, 13-15 rue du Pont des Halles, 94526 Rungis, France
123

90
J Am Oil Chem Soc (2012) 89:89–96
acyl migration occurring during deacylation. Other methods
have been developed with the use of 13C-NMR or LC–MS
equipment [6–8]. Although these methods are efficient they
impose the use of expensive tools which are not readily
available. Finally, probably the most widely used method
for the determination of triacylglycerol regiodistribution is
an enzymatic one, involving the use of pancreatic lipase [9].
This enzyme being known as strictly 1,3 regioselective, its
use in such determinations consists of the partial hydrolysis
of TAG substrate leading to the formation of acylglycerols
and in particular b-monoacylglycerols (b-MAG) which are
representatives of the native TAG. Then, the analysis of this
b-MAG allows us to determine the FA composition of the
internal position of native TAG. However, due to the pan-
creatic lipase selectivity for some FA, this method shows
some restrictions. Actually, short-chain FA are more readily
hydrolyzed by pancreatic lipase than longer chain ones,
what leads to the formation of non representative b-MAG
and consequently incorrect estimation of FA positional
distribution. Thus, this method cannot be used for the
analysis of fat containing a high content of medium and
short-chain FA such as milk fat.
sodium borohydride (NaBH4) and solvents were purchased
from Sigma-Aldrich (St Quentin Fallavier, France). 1,3-
Dihydroxyacetone was purchased from VWR (Fontenay
sous Bois, France). Tris buffer 1 M was prepared from tris-
(hydroxymethyl)aminomethane and its pH was adjusted
by HCl 2 M. Lipase from Rhizopus oryzae (55 U/mg, 1 U
corresponds to the amount of enzyme which liberates
1 lmol fatty acid from triacylglycerols per minute at pH
7.2 and 37 °C, olive oil as substrate) and lipase from
porcine pancreas (Type II, 100–400 U/mg protein, one unit
will hydrolyze 1.0 microequivalent of fatty acid from olive
oil in 1 h at pH 7.7 at 37 °C.) were purchased from Sigma-
Aldrich.
Analysis
Thin Layer Chromatography (TLC) Densitometry
Analytical TLC was performed on pre-coated silica-gel
TLC plates (TLC Silica gel 60, 10 9 20 cm, Merck,
Fontenay sous bois, France). After development in a
suitable elution solvent (as detailed in each section), TLC-
plates were analyzed by densitometry. Densitometric
evaluations were performed with a TLC-scanner 3 (CA-
MAG, Switzerland) in absorbance mode at k = 190 nm.
Data were processed with the software winCATS.
Consequently, although this enzymatic method is satis-
factory due to its relative simplicity, there is still a need in
identifying new lipases with limited acyl selectivity that
could be an alternative to pancreatic lipase. Indeed, the
development of a new enzymatic method involving a lipase
which acts similarly on short chain or long chain fatty acids
would broaden the range of application of such a method,
allowing its use on substrates containing short aliphatic
chains like hard stock used in margarine formulation.
Recently, two 1,3 regiospecific lipase, namely Rhizopus
oryzae [10, 11] and Rhizopus arrhizus [12, 13] have been
proposed, but these studies were limited to TAG containing
medium and long chain lengths. However, in order to
evaluate the possibility of one of these lipases as an
alternative to pancreatic lipase, it is necessary to assess its
activity toward a wider range of substrates and especially
those with short chains fatty acids down to eight carbons.
Accordingly, in the present paper we have studied the
behavior of Rhizopus oryzae lipase on the hydrolysis of
model structured triacylglycerols containing caprylic acid
in the central sn-2 position and compare its 1,3 regiose-
lectivity and typoselectivity to the one of pancreatic lipase.
Flash-Chromatography
Purification steps were performed by flash liquid chroma-
tography using a CombiFlash Companion (Teledyne Isco,
USA) and RediSep prepacked columns. Column size, flow
rate, solvents and gradient used are detailed in each puri-
fication section.
1H- and 13C-NMR Spectroscopy
1H-NMR spectra were recorded on a Bruker Avance DRX
400 spectrometer at 400 MHz in deuterochloroform
(CDCl3) as solvent. Chemical shifts (d) are given in parts
per million (ppm) and the coupling constant (J) in Hertz
(Hz). Each resonance is shown according to the following
convention: chemical shift (d ppm) (multiplicity, coupling
constant (Hz), number of hydrogens, assignment).
13C-NMR spectra were recorded on the same spectrometer
at 100 MHz in CDCl3 as solvent. For each 13C signal, the
number of carbon nuclei is indicated in parentheses when
there is more than one carbon responsible for the peak.
Materials and Methods
Chemicals and Lipases
Synthesis of Model TAG
Caprylic acid (C99%), oleic acid ([90%), triolein (C99%),
tricaprylin (C99%), trilaurin (C99%), N,N0-dicyclohexyl-
carbodiimide (DCC), 4-dimethylaminopyridine (DMAP),
Pure 1,3-dioleoyl-2-capryloyl-sn-glycerol (OCO) was
chemically synthesized in 3 steps.
123

J Am Oil Chem Soc (2012) 89:89–96
91
Synthesis of 1,3-Dioleoylpropane-2-one
load cartridge. Elution was then carried out on a 40 g sil-
ica-gel column with a gradient of chloroform to chloro-
form/acetone (95:5, v/v) achieved in 45 min at a flow rate
of 10 ml/min. Fractions of 3 ml were collected and ana-
lyzed by TLC-densitometry as described above. Fractions
of pure 1,3-dioleate (Rf = 0.34) were combined and the
solvent evaporated. Final yield = 85%.
Oleic acid (0.05 mol) and 4-DMAP (0.05 mol) were added
to a solution of 1,3-dihydroxyacetone (0.025 mol) in 50 ml
of chloroform. A solution of DCC (0.05 mol) in 25 ml of
chloroform was then added dropwise under stirring at room
temperature. Formation of 1,3-dioleoylpropane-2-one was
monitored by TLC in hexane/ethyl acetate (80:20, v/v).
After overnight reaction, the precipitated dicyclohexylurea
was removed by filtration and the solvent eliminated under
vacuum (rotavapor).
Synthesis of Pure 1,3-Dioleoyl-2-Capryloyl-sn-Glycerol
(OCO)
The 1,3-dioleoylpropane-2-one was then purified by
flash chromatography according to the following proce-
dure: 4 g of filtrate were adsorbed on silica and placed in a
solid load cartridge. Elution was then carried out on a 80 g
silica-gel column with a gradient of hexane to hexane/ethyl
acetate (70:30, v/v) (initial 100% hexane was held for
15 min and increase of ethyl acetate content was achieved
in 40 min) at a flow rate of 30 ml/min. Separation was
followed by UV-detection at 200 nm and fractions of
10 ml were collected and analyzed by TLC-densitometry.
TLC plates were developed in hexane/ethyl acetate
80:20 (v/v) and scanned at k = 190 nm. Fractions of pure
1,3-dioleoylpropane-2-one (Rf = 0.55) were combined and
the solvent evaporated under vacuum.
Caprylic acid (10 mmol) in 4 ml of chloroform and
4-DMAP (3 mmol) were added to a solution of the
previously synthesized 1,3-dioleate (3 mmol) in 30 ml of
chloroform. The reaction mixture was heated at 30 °C
under magnetic stirring and a solution of DCC (10 mmol)
in 20 ml of chloroform was then added in small portions
(*30 min). Formation of triacylglycerols (TAG) was
followed by GC on a Thermo instrument (Courtaboeuf,
France), GC 8000 series, equipped with a flame ioniza-
tion detector and Supelco’s Equity 1 capillary column
(15 m 9 0.32 mm, film thickness 0.1 lm). Helium was
used as carrier gas at a flow rate of 5 ml/min. The
operating conditions were: 1 ll injection on-column,
oven temperature programmed from 80 to 330 °C at
10 °C/min with a final hold of 10 min, detector at
350 °C.
Synthesis of 1,3-Dioleate
On completion of the reaction, the precipitated dicy-
clohexylurea was removed by filtration and the solvent
evaporated. Formation of 1,2(2,3)-dioleate during the
reaction was monitored by TLC-densitometry as described
above.
1,3-dioleoylpropane-2-one (3.5 mmol) was dissolved in
THF (45 ml) and distilled water (3 ml). This heteroge-
neous solution was cooled to 5 °C in a water bath under
magnetic stirring, and sodium borohydride (*8 mmol)
was carefully added. Reaction was followed by TLC in
solvent system hexane/diethyl ether/acetic acid 70:30:1
(v/v/v). After 30 min, 1,3-dioleoylpropane-2-one was
totally reduced in 1,3-dioleate and excess borohydride was
neutralized by dropwise addition of glacial acetic acid
(1 ml). The solution was diluted in 100 ml of chloroform,
washed with water (*100 ml), aqueous sodium bicar-
bonate (*100 ml) and water (*100 ml), dried over
anhydrous sodium sulfate, filtered and the solvent evapo-
rated under vacuum.
The TAG contained in the filtrate was finally purified by
flash-chromatography according to the following proce-
dure: 3.7 g of filtrate was adsorbed on silica and placed in a
solid load cartridge. Elution was then carried out on a 12 g
silica-gel column with a gradient of hexane to hexane/
diethyl ether (70:30, v/v) achieved in 60 min at a flow rate
of 10 ml/min. The effluent was monitored by UV-absorp-
tion at k = 200 nm. Fractions of 5 ml were collected and
analyzed by TLC-densitometry in hexane/diethyl ether/
acetic acid (70/30/1, v/v/v). Fractions of pure TAG
(Rf & 0.55) were collected and combined. Final
yield = 50%. 1H NMR d 5.40–5.30 (m, 4H, =C–H),
5.30–5.27 (m, 1H, CH2–CH–CH2), 4.33–4.29 (dd, J =
11.9 Hz, 2H, CH2–CH–CH2), 4.19–4.14 (dd, J = 11.9 Hz,
2H, CH2–CH–CH2), 2.35–2.31 (m, 6H, CH2–CO),
2.08–2.00 (m, 8H, CH2–CH=CH), 1.64–1.61 (m, 6H, CH2–
CH2–CO), 1.32–1.29 (m, 48H, –CH2), 0.92–0.88 (m, 9H,
–CH3). 13C NMR d 174.65 (2, aC=O), 174.26 (1, bC=O),
131.62–129.31 (4, C=C), 70.29 (1, CH), 63.52 (2, OCH2),
36.08–23.99 (34, CH2), 15.52–15.47 (3, CH3).
Due to acyl migration, the unwanted isomer (1,2(2,3)-
dioleate) can also be formed during the reaction. This
isomerization was monitored by TLC-densitometry: TLC
plates were developed in hexane/diethyl ether/acetic acid
(70:30:1, v/v/v) and scanned at k = 190 nm. 1,2(2,3)-di-
oleate and 1,3-dioleate had a Rf of 0.25 and 0.34,
respectively.
Purification of 1,3-dioleate was achieved by flash
chromatography according to the following procedure: the
above dioleate was adsorbed on silica and placed in a solid
123

92
J Am Oil Chem Soc (2012) 89:89–96
Lipase Specificities Measurement
prepared from tricaprylin (329.4 mg, 0.7 mmol), triolein
(619.7 mg, 0.7 mmol) and hexane to obtain a 100 mg/ml
solution. After vigorous shaking, 1 ml of the blend was
taken, dried under nitrogen and used as substrate for sub-
sequent hydrolysis. A second blend was prepared in the
same conditions from trilaurin (319.7 mg, 0.5 mmol) and
triolein (443.7 mg, 0.5 mmol). The hydrolysis procedure
was the same than described above but adding 0.5 ml of
lipase solution (20 mg/ml).
1,3-Regioselectivity
Twenty milligrams of pure OCO in 0.1 ml of hexane and
2 ml of 1 M tris-buffer, pH 7 were heated at 40 °C for
5 min. 100 ll of the Rhizopus oryzae (RO) lipase solution
(20 mg/ml) were added and the mixture was manually
shaken at 40 °C for 1 min, then shaken vigorously for
1 min, followed by shaking at 40 °C for 1 min. After 3 min
of hydrolysis, 1 ml of 2 M HCl solution was added to stop
the reaction. Then 2 ml of diethyl ether was added and the
medium was vigorously shaken to extract the reaction
products. After centrifugation at 790 g for 5 min, the upper
layer of the organic phase was collected and separated by
semi-preparative TLC: 250 ll of the organic phase were
spotted on a silica gel plate which was then developed with
a hexane/diethyl ether/acetic acid solution (50:50:1 v/v/v).
After a short drying time, a part of the plate was cut and the
lipid spots were detected by immersion in a solution of
saturated copper sulfate/phosphoric acid (50:50, v/v) and
heating at 180 °C for 10 min. The b-MAG band
(Rf & 0.08) was scraped off from the plate and its FA
composition was determined by GC after conversion into
methyl esters according to the following procedure: 3 ml
sodium methylate solution containing phenolphthalein
were added directly to the silica gel. The reaction medium
was refluxed for 10 min. Then 3 mL chlorohydric metha-
nol (acetyl chloride in methanol, 1 M) were added until
phenolphthalein discoloration and the mixture was refluxed
again for 10 min and then cooled to room temperature.
Finally, 3 ml hexane and 10 ml water were added and the
organic phase containing fatty acids methyl esters (FAME)
was recovered for subsequent analysis. FAME were ana-
lyzed by GC (Agilent 6890 series, Bios Analytique,
France) equipped with a flame ionization detector, and a
Supelcowax capillary column (SGE, Courtaboeuf, France)
with the following characteristics: 30 m 9 0.32 mm; film
thickness 0.25 lm. Oven was heated from 100 to 225 °C at
5 °C/min and held at 225 °C for 5 min. Detector and
injector temperatures were 250 and 270 °C, respectively,
and helium was used at a flow rate of 1 ml/min. Peak area
percentages obtained with the integrator were divided by
the molecular weight of individual FAME to yield mole
percent of fatty acids. Experiments were performed in
triplicate, and results given correspond to the mean values
of three determinations.
Hydrolysis by Pancreatic Lipase
Partial hydrolysis using pancreatic lipase was achieved on
OCO and on the equimolar blends of homogeneous TAG as
described in the official method for regiodistribution
analysis [9].
Results and Discussion
Synthesis of Model TAG
Evaluation of Rhizopus oryzae lipase 1,3-selectivity was
based on the hydrolysis of pure 1,3-dioleoyl-2-capryloyl-
sn-glycerol (OCO) (Fig. 1). FA composition and regiodis-
tribution of this substrate has to be perfectly known in
order to predict hydrolysis products and conclude regarding
the lipase selectivity. Various strategies have been descri-
bed for the synthesis of highly pure triacylglycerols. These
methods have been recently reviewed by Wijesundera
et al. [14, 15] and generally require numerous chemical or
enzymatic steps and tedious subsequent purification. The
main difficulty to overcome during such synthesis is to
avoid acyl migration phenomena which can easily occur
depending on many parameters (e.g., alkalinity of the
medium, temperature) [16–18]. Therefore, we aimed our
strategy at the synthesis of pure OCO via a 1,3-dioleate
intermediate of high purity and minimum formation of
1(3),2-DAG isomers. Actually, long-chain DAGs typically
display a 1:2 ratio of 1,2(2,3)-DAG to 1,3-DAG at equi-
librium [18]. Purification of 1,3-dioleate is required to
O
O
O
O
Typoselectivity
O
O
Chain length selectivity of RO lipase was assessed by
investigating partial hydrolysis of equimolar blends of
pure homogeneous triacylglycerols. The first blend was
Fig. 1 Structured triacylglycerol OCO. O is oleic acid (C18:1) and C
is caprylic acid (C8:0)
123

J Am Oil Chem Soc (2012) 89:89–96
93
Fig. 2 Overall synthesis
synthesis of OCO
1,3-dioleate
oleic
oleic
OH
OH
oleic
oleic
oleic
oleic acid
NaBH4
caprylic acid
DCC/DMAP
O
O
OH
caprylic
DCC/DMAP
oleic
isomerization
oleic
oleic
caprylic acid
DCC/DMAP
oleic
oleic
OH
caprylic
1,2(2,3)-dioleate
synthesis of OOC
avoid the formation of the unwanted isomer 1,2-dioleoyl-
3-octanoyl-sn-glycerol (OOC) from 1,2(2,3)-dioleate.
Accordingly, 1,3-dioleoyl-2-capryloyl-sn-glycerol (OCO)
was synthesized in three steps as displayed in Fig. 2. The
first step, adapted from the method described by Kodali
et al. [19], consisted of the synthesis of 1,3-dioleoylpro-
pane-2-one by acylation of dihydroxyacetone. This had the
advantage of starting from a cheap and easily available
material. 1,3-dioleoylpropane-2-one was obtained in high
purity after flash chromatography and used in the second
step, as described by Bentley et al. [20]. This step consisted
in the reduction of the central keto group of 1,3-dioleoyl-
propane-2-one by action of sodium borohydride to form
1,3-dioleate. At this stage, acyl migration occurred and led
to the formation of the isomer 1,2(2,3)-dioleate. Before
purification, the ratio 1,3-/1,2(2,3)-dioleate obtained was
86/14 (Table 1). After purification by flash-chromatogra-
phy, no 1,2(2,3)-dioleate was detected in the final product,
neither by densitometric evaluation at k = 190 nm nor by
detection after immersion in a solution of saturated copper
sulfate/phosphoric acid (50:50, v/v) and heating at 180 °C
for 10 min. The final step, adapted from the method
described by Kodali et al. [19] and Adlof et al.[21], con-
sisted in the synthesis of OCO by acylation of 1,3-dioleate.
Isomerization was also monitored during this step. No
1,2(2,3)-dioleate was detected, what allowed us to expect
high purity for OCO.
Table 1 Purity of the intermediate and final products of OCO
synthesis
Synthesis step
Purity of the isomer
Purity evaluation
method
Synthesis of 1,3-dioleate
Before purification
After purification
86% of 1,3-DAG
TLC-densitometry
TLC-densitometry
No 1,2(2,3)-DAG
detected
Synthesis of OCO
[99.2%
Pancreatic lipase
method
Table 2 Fatty acids composition of MAG resulting from hydrolysis
of OCO by RO and pancreatic lipases
Mol (%)
RO lipase
Pancreatic lipase
C8:0
99.6 0.2
0.4 0.2
99.2 0.6
0.8 0.6
C18:1
last steps of synthesis. It has to be reminded that hydrolysis
by pancreatic lipase is influenced by FA chain length due to
lipase typoselectivity. This selectivity concerns short
and medium chain FA which means that in the case of
OCO/OOC isomers, and taking also the strict 1,3-selec-
tivity into account, pancreatic lipase would tend to
hydrolyze caprylic acid of the external positions faster than
oleic acid. Thus, if caprylic acid occurs on external posi-
tions, corresponding to the presence of the unwanted iso-
mer OOC, the typoselectivity will induce an increase of the
formation of monoolein. Consequently, typoselectivity of
pancreatic lipase can lead to an overestimation of the oleic
content of MAG formed, and, as a result, to an overesti-
mation of the OOC content. This means that the purity of
the OCO isomer estimated from the FA composition of
MAG resulting from hydrolysis by pancreatic lipase will be
Regiodistribution of the final product was analyzed by
the pancreatic lipase method. Due to the strict 1,3-selec-
tivity of this lipase, partial hydrolysis led to the formation
of b-MAG. The analysis of their FA composition showed
that MAG resulting from the hydrolysis of the final product
were almost exclusively made of monocaprylin. Actually,
recovered MAG were constituted at 99.2% of caprylic acid
(Table 2). Thus, the internal position was contaminated by
only 0.8% of oleic acid due to acyl migration during the
123

94
J Am Oil Chem Soc (2012) 89:89–96
underestimated. Consequently, according to these consid-
erations and the results obtained, we can say that purity of
the 1,3-dioleoyl-2-capryloyl-sn-glycerol synthesized was
higher than 99.2% (Table 1).
of MAG released from hydrolysis by the two lipases are
shown in Fig. 3. These results revealed that MAG formed
by hydrolysis of CCC/OOO by RO lipase had the same FA
composition than the initial blend showing that the lipase
hydrolyzed the TAG without any FA chain length prefer-
ence. Considering results obtained for hydrolysis of OCO,
it also showed that, in this hydrolysis conditions, RO lipase
keeps its 1,3-selectivity even in presence of short-chain
fatty acids in central position of the TAG. On the contrary,
MAG formed by hydrolysis of CCC/OOO by pancreatic
lipase consisted of more than 80% of caprylic acid (Fig. 3)
meaning that pancreatic lipase hydrolyzed tricaprylin more
preferentially than triolein, due to a typoselectivity toward
caprylic acid or at least short chain FA typoselectivity.
FA compositions of the initial LLL/OOO blend and of
MAG released from hydrolysis by the two lipases are
Lipase Selectivity Evaluation
The official method for regiodistribution analysis involves
the action of pancreatic lipase. This method especially uses
the strict 1,3-selectivity of the lipase under partial hydro-
lysis conditions which leads to the formation of b-MAG
representative of native TAG. Subsequent analysis of this
b-MAG allows direct determination of the FA composition
of internal position. However, due to the specificity of
pancreatic lipase, this method cannot be used for the
analysis of fats containing a significant quantity of FA with
12 carbons or less [4, 9]. Actually, pancreatic lipase tends
to hydrolyze these FA faster than longer chain ones leading
to the formation of unrepresentative MAG. Consequently,
overcoming these restrictions by finding an alternative to
pancreatic lipase for regiodistribution analysis amounts
to finding a lipase both 1,3-selective and with limited acyl
selectivity toward FA chainlength.
100
80
C18:1
C8:0
Our main objective was to estimate whether Rhizopus
oryzae lipase exhibits a strict 1,3-selectivity when cata-
lyzing the hydrolysis of TAG containing C8 carbon chain
length fatty acids. For this, 1,3-dioleoyl-2-capryloyl-
sn-glycerol (OCO) appeared to be particularly appropriate
due to the location of this fatty acid at the central sn-2
position of this structured substrate. Indeed, investigating
the partial hydrolysis of OCO can give valuable informa-
tion regarding the conservation of the 1,3-regioselectivity
of the tested lipase and its competition with any potential
typoselectivity (acyl selectivity for C8 FA). The conditions
of hydrolysis were adapted from the method described by
Kosugi et al. [10] and a strict regio preference for the two
external position of the triglycerides backbone of OCO
would result in the release of only oleic acid as free fatty
acid and concomitantly the appearance of MAG exclu-
sively made of caprylic acid. Results obtained for FA
composition of the newly formed MAG showed that they
were almost exclusively made of monocaprylin as caprylic
acid accounting for 99.6% of the fatty acid composition of
recovered MAG (Table 2). This result showed that Rhizo-
pus oryzae lipase maintains its strict 1,3-regioselectivity
even though C8 fatty acids are present in the triglyceridic
structure of the substrate.
60
81
40
20
0
43
41
initial
blend
MAG
pancreatic
lipase
MAG
RO lipase
Fig. 3 FA composition of initial blend CCC/OOO and of MAG
resulting from hydrolysis of the blend by RO and pancreatic lipases
100
80
C18:1
60
C12:0
40
73
65
47
20
For typoselectivity evaluation, partial hydrolysis by
Rhizopus oryzae lipase was carried out on equimolar blends
of tricaprylin/triolein (CCC/OOO) and trilaurin/triolein
(LLL/OOO) and results were compared with those
obtained for hydrolysis of the same blends by pancreatic
lipase. FA compositions of the initial CCC/OOO blend and
0
MAG
pancreatic
lipase
initial
blend
MAG
RO lipase
Fig. 4 FA composition of initial blend LLL/OOO and of MAG
resulting from hydrolysis of the blend by RO and pancreatic lipases
123

J Am Oil Chem Soc (2012) 89:89–96
95
Acknowledgment This study was performed in the framework of a
Ph.D. study with financial support from the St Hubert company in
France.
shown in Fig. 4. These results show that both lipases
hydrolyzed trilaurin more preferentially than triolein. For
pancreatic lipase, it confirms its typoselectivity for medium
chain FA. Concerning RO lipase results tend to show a
typoselectivity for lauric acid with 65% of the newly
formed MAG being made of monolaurin. However, we
cannot exclude the hypothesis that this preferential
hydrolysis is due to a modification of the emulsion. Actu-
ally, hydrolysis forms MAG and DAG which are known for
their emulsifying properties. The hydrophilic/lipophilic
balance of partial acylglycerols is function of the chain
length of FA esterified on the glycerol [22]. MAG and
DAG resulting from the hydrolysis of trilaurin (12-MAG
and 12-DAG) have a better hydrophilic/hydrophobic bal-
ance than MAG and DAG resulting from hydrolysis of
triolein (18-MAG and 18-DAG). Thus, whereas 18-MAG
and 18-DAG would tend to form micelles or stay in the oil
phase because of their high hydrophobicity, 12-MAG and
12-DAG would tend to emulsify TAG droplets. By placing
themselves at the oil/water interface where the enzymatic
reaction takes place [23], 12-DAG are more readily
hydrolyzed, what leads to an increase in 12-MAG content.
However, for hydrolysis of both blends, the RO lipase
was less influenced by fatty acid chain length than pan-
creatic lipase. It means that MAG resulting from the
hydrolysis by RO lipase are more representative of the
initial TAG than MAG from hydrolysis by pancreatic
lipase. Considering also its 1,3-specificity, the RO lipase
turns out to be a better candidate for the regiodistribution
analysis of fat containing medium and short chain FA.
References
1. Xu X (2000) Production of specific-structured triacylglycerols
by lipase-catalyzed reactions: a review. Eur J Lipid Sci Technol
102:287–303
2. Mu H, Porsgaard T (2005) The metabolism of structured tria-
cylglycerols. Prog Lipid Res 44:430–448
3. Brockerhoff H (1967) Stereospecific analysis of triglycerides:
an alternative method. J Lipid Res 8:167–169
4. Turon F (2003) Clarification on the analysis of the regiodistri-
bution in fatty acids of triacylglycerols. Ocl-Oleag Corps Gras
Lipides 10:144–149
5. Turon F, Bachain P, Caro Y, Pina M, Graille J (2002) A direct
method for regiospecific analysis of TAG using alpha-MAG.
Lipids 37:817–821
6. Mottram HR (2005) Regiospecific analysis of triacylglycerols
using high performance liquid chromatography/atmospheric
pressure chemical ionization mass spectrometry. In: Byrdwell
WC (ed) Modern methods for lipid analysis by liquid chroma-
tography/mass spectrometry and related techniques. AOCS Press,
Champaign, pp 276–297
7. Vlahov G (1998) Regiospecific analysis of natural mixtures of
triglycerides using quantitative C-13 nuclear magnetic resonance
of acyl chain carbonyl carbons. Magn Reson Chem 36:359–362
8. Blaise P, Wolff R, Farines M (1997) Regiospecific study of
vegetable oil triacylglycerols by chemical cleavage and high-
resolution NMR C-13. Ocl-Oleag Corps Gras Lipides 4:135–141
9. AFNOR (2000) Determination of the composition of fatty acids
in the 2-position of the triglycerides molecules. In: AFNOR (ed)
´
´ ´
Corps gras, graines oleagineuses, produits derives Tome 1: corps
´
´
gras et produits derives, pp 189–202
10. Kosugi Y, Oshima A, Koike S, Fukatsu M, Minami K, Miyake Y,
Masui K (2002) Determination of fatty acid composition at sn-2
acyl position in triacylglycerol by capillary gas chromatography
with lipase from Rhizopus delemar. J Oleo Sci 51:599–605
11. Kosugi Y, Oshima A, Koike S, Fukatsu M, Minami K, Miyake Y,
Masui K (2004) Use of Rhizopus delemar lipase as compared
with other lipases for determination of sn-2 fatty acids in triac-
ylglycerol (vol 81, pg 235, 2004). J Am Oil Chem Soc 81:523
12. Amate L, Ramirez M, Gil A (1999) Positional analysis of tri-
glycerides and phospholipids rich in long-chain polyunsaturated
fatty acids. Lipids 34:865–871
13. Arcos JA, Garcia HS, Hill CG (2000) Regioselective analysis of
the fatty acid composition of triacylglycerols with conventional
high-performance liquid chromatography. J Am Oil Chem Soc
77:507–512
14. Wijesundera C (2005) Synthesis of regioisomerically pure tria-
cylglycerols containing n-3 very long-chain polyunsaturated fatty
acids. Eur J Lipid Sci Technol 107:824–832
15. Haraldsson GG, Halldorsson A, Kulas E (2000) Chemoenzymatic
synthesis of structured triacylglycerols containing eicosapentae-
noic and docosahexaenoic acids. J Am Oil Chem Soc 77:1139–
1145
16. Halldorsson A, Magnusson CD, Haraldsson GG (2003) Chemo-
enzymatic synthesis of structured triacylglycerols by highly
regioselective acylation. Tetrahedron 59:9101–9109
17. Li W, Du W, Li Q, Sun T, Liu D (2010) Study on acyl migration
kinetics of partial glycerides: dependence on temperature and
water activity. J Mol Catal B: Enzym 63:17–22
Conclusion
Synthesis of the structured triacylglycerol 1,3-dioleoyl-2-
capryloyl-sn-glycerol (OCO) was achieved by action of
DCC/DMAP coupling agent and purification steps by flash-
chromatography. The final product showed a very high
purity with less than 1% of the unwanted isomer OOC.
This structured TAG was used as substrate to evaluate
the regioselectivity of the Rhizopus oryzae lipase. Results
showed a strict 1,3-selectivity of the lipase even in pres-
ence of short chain FA in central position. Typoselectivity
of the lipase was evaluated by hydrolyzing equimolar
blends of homogeneous TAG containing 8 or 12 carbons
chain lengths. It turned out that the RO lipase was less
influenced by FA chain length than the pancreatic lipase
generally used for regiodistribution analysis.
Considering these results of lipase selectivity evalua-
tion, it appeared that Rhizopus oryzae lipase should be
preferentially used for the determination of regiodistribu-
tion in TAG containing short and medium chain FA.
123

96
J Am Oil Chem Soc (2012) 89:89–96
18. Laszlo JA, Compton DL, Vermillion KE (2008) Acyl migration
kinetics of vegetable oil 1,2-diacylglycerols. J Am Oil Chem Soc
85:307–312
21. Adlof RO, List GR (2008) Synthesis and physical properties of
symmetrical and non-symmetrical triacylglycerols containing two
palmitic fatty acids. J Am Oil Chem Soc 85:99–104
19. Kodali DR, Atkinson D, Redgrave TG, Small DM (1987)
Structure and polymorphism of 18-carbon fatty acyl triacylgly-
cerols: effect of unsaturation and substitution in the 2-position.
J Lipid Res 28:403–413
20. Bentley PH, McCrae W (2002) Synthesis of symmetrical
1,3-diglycerides. J Org Chem 35:2082–2083
22. Gunstone F, Harwood J, Dijkstra A (2007) The lipid handbook
with CD-ROM, 3rd edn. CRC Press, Boca Raton
23. Fickers P, Destain J, Thonart P (2008) Lipases are atypical
hydrolases: Principal characteristics and applications. Biotechnol
Agron Soc Environ 12:119–130
123