Identification and quantification of feruloylated mono-, Di-, and triacylglycerols from vegetable oils

Article abstract of DOI:10.1007/s11746-006-5010-5

The use of HPLC-MS to separate and identify the feruloylated acylglycerols formed during the transesterification of ethyl ferulate with TAG was examined. Novozym 435 (Candida antarctica lipase B)-catalyzed transesterifications of ethyl ferulate and soybea

Full text of DOI:10.1007/s11746-006-5010-5

Identification and Quantification of Feruloylated
Mono-, Di-, and Triacylglycerols from Vegetable Oils
David L. Compton*, Joseph A. Laszlo, and Mark A. Berhow
New Crops and Processing Technology Research Unit, National Center for Agricultural
Utilization Research, ARS, USDA, Peoria, Illinois 61604
ABSTRACT: The use of HPLC–MS to separate and identify the tion of fats and oils, including regular-phase (5), reversed-
feruloylated acylglycerols formed during the transesterification of
ethyl ferulate with TAG was examined. Novozym 435 (Candida
antarctica lipase B)-catalyzed transesterifications of ethyl ferulate
and soybean oil resulted in a mixture of feruloylated MAG, DAG,
and TAG and diferuloylated DAG and TAG. These feruloylated
acylglycerols have recently garnered much interest as cosmeceu-
tical ingredients. The ratio of the various feruloylated acylglyc-
erol species in the resultant oils is presumed to affect the oil’s cos-
metic efficacy as well as its physical (formulation) properties.
Thus, it was desirable to develop an analytical method to sepa-
rate, identify, and quantify the individual feruloylated acylglyc-
phase (6,7), argentation (silver ion) (8,9), 2-D silver-ion re-
versed-phase (10), and size-exclusion (11) chromatography. An
examination of various HPLC methods for resolving acyl lipids
recently has been published (12). These HPLC methods are
used to resolve TAG species based on fatty acid composition,
positional isomers, and degree of desaturation. The biocatalytic
and chemical conversion of TAG to structured lipids and fatty
acid esters results in intermediate DAG and MAG as well as
FFA. HPLC methods have been developed to resolve the DAG
and MAG and elucidate the positional isomers of these inter-
erols to determine their relative ratios. The feruloylated acylglyc- mediate glycerols (13–15). A few of these HPLC methods have
erols were successfully separated and identified by HPLC–MS
using a phenyl-hexyl reversed-phase column developed with a
water/methanol/1-butanol gradient. The chromatograms of the
feruloylated acylglycerols from soybean oil were convoluted by
myriad fatty acids; therefore, feruloylated acylglycerols from tri-
olein were studied as a model reaction. Hydrolysis of the feruloy-
lated acylglycerols from triolein catalyzed by Lipase PS-C
“Amano” I (Burkholderia cepacia), which showed no hydrolysis
reactivity toward ethyl ferulate, allowed for the chromatographic
assignment of the feruloyl acylglycerol positional isomers.
been reviewed in detail elsewhere (16).
Partial substitution of fatty acid groups in vegetable oils by
aryl groups presents new analytical challenges. HPLC methods
that have previously been developed for the resolution of phe-
nolic lipids have focused on catecholic (e.g., urushiol, anac-
ardic, ginkgolic), resorcinolic, and hydroquinonic lipids, which
are more structurally similar to FFA (17,18). The present work
establishes an HPLC method to separate and identify various
feruloyl (phenolic)-substituted MAG, DAG, and TAG.
A reversed-phase HPLC method using a water/methanol/1-
butanol gradient and a phenyl-hexyl column was used to eluci-
date the molecular species derived from enzymatically feruloy-
lated vegetable oils. The lipase-catalyzed transesterification of
ethyl ferulate (EF) and soybean oil (SBO) resulted in a mixture
of feruloylated acylglycerol species (19,20). The UV absorp-
tion and antioxidant properties of the feruloylated acylglycerols
have recently led to interest in them as all-natural ingredients
for use in daily-wear cosmetics and skin-care products. Each
of the feruloylated acylglycerol molecular species is presumed
to have intrinsic chemical and physical properties that will af-
fect the overall efficacy and formulation capabilities of the bulk
oil. Thus, it was desirable to develop an HPLC method to sep-
arate, identify, and quantify the various feruloylated acylglyc-
erol species. This method will guide the optimization of reac-
tion parameters such as reactant ratio and reaction time to con-
trol the relative ratios of the feruloylated species in the resultant
oil, which influence the optimization of the ingredient’s effi-
cacy and formulation properties.
Paper no. J11296 in JAOCS 83, 753–758 (September 2006).
KEY WORDS: Ferulic acid, feruloylated lipids, HPLC–MS, li-
pase, lipids, transesterification.
Vegetable oils have been biocatalytically and chemically con-
verted to commercially successful structured lipids for food
and skin-care applications (1). Vegetable oils have also been
used as feedstocks for large-scale industrial products such as
surfactants, paints, lubricants, and biodiesel (2–4). These bio-
catalytic and chemical processes are rarely quantitative and
often result in MAG and DAG intermediates and by-products
as well as free glycerol, FFA, and unreacted TAG. These inter-
mediates and by-products are often unwanted in the final prod-
uct; thus, methods have been developed to identify and moni-
tor these contaminants during synthesis and purification.
A convenient, versatile method for separation and identifi-
cation of vegetable lipids and their intermediate glycerols is
HPLC incorporating UV detection, ELSD, and MS. A variety
of HPLC methods have been developed to optimize the resolu-
EXPERIMENTAL PROCEDURES
*To whom correspondence should be addressed at New Crops and Process-
ing Technology Research Unit, NCAUR, ARS, USDA, 1815 N. University
St., Peoria, IL 61604. E-mail: comptond@ncaur.usda.gov
Materials. EF (ethyl 4-hydroxy-3-methoxy cinnamate) was
purchased from Shanghai OSD (Shanghai, China). Novozym
Copyright © 2006 by AOCS Press
753
JAOCS, Vol. 83, no. 9 (2006)

754
D.L. COMPTON ET AL.
435 (Candida antarctica lipase B immobilized on acrylic
beads) was purchased from Novozymes North America
(Franklinton, NC). Lipase PS-C “Amano” I (Burkholderia
cepacia immobilized on ceramic particles) was graciously do-
nated by Amano Enzyme USA Co. (Elgin, IL). Triolein was
purchased from Nu-Chek-Prep (Elysian, MN). SBO was pur-
chased at a local grocery. All other reagents were from Sigma-
Aldrich (St. Louis, MO). Solvents and glycerol (<0.1% w/w
water) were spectroscopic grade.
Synthesis of feruloylated mono- and dioleoylglycerols. Vary-
ing amounts of glycerol (0.0 to 0.368 g, 0.0 to 4.2 mmol) were
blended with 200 mg of silica gel in 50-mL Schlenk flasks. Tri-
olein (3.54 g, 4.0 mmol) and Novozym 435 (0.221 g) were
added to the Schlenk flasks and the mixtures were degassed
under vacuum for 30 min. The evacuated flasks were shaken
on an orbital shaker (200 rpm) at 60°C for 24 h. EF (0.888g,
4.0 mmol) was added to the flasks, and the flasks were evacu-
ated for 5 min. The reaction mixtures were allowed to shake
(200 rpm) at 60°C for 120 h. Aliquots (100 mL) were removed
at timed intervals for analytical HPLC and UV spectroscopy
(325 nm). The flasks were evacuated for 5 min after each sam-
pling. The feruloylated reaction products (Fig. 1) were identi-
fied by HPLC-MS (see below).
FIG. 1. The feruloylated species obtained from Novozym 435-catalyzed
transesterification of triolein or soybean oil (SBO) with ethyl ferulate (EF).
The structures shown do not denote all possible sn-positional isomers. F-
MAG, feruloylated MAG; F-DAG, feruloylated DAG; F-TAG, feruloylated
TAG; FMOG, feruloyl monooleoylglycerol; FDOG, feruloyl dioleoylglyc-
erol; F2-DAG, diferuloylated DAG; F2-TAG, diferuloylated TAG; F2MOG,
diferuloyl monooleoylglycerol.
Deacylation of feruloylated mono- and dioleoylglycerols.
After reaching equilibrium (120 h), the feruloylated oleoylglyc-
erol reaction mixtures were extracted with 10 mL of 2-methyl-
2-propanol. The solutions were filtered to remove the analyzed by analytical HPLC and was found to be pure. Yield:
Novozym 435 and silica gel. The solids were rinsed with two 3.69 g. 1H NMR (d6-acetone): δ (ppm) 7.60 (1 H, d, J = 16.1
2-mL portions of 2-methyl-2-propanol (t-BuOH). The com- Hz, H-8), 7.31 (1 H, s, H-2), 7.12 (1 H, d, J = 8.2 Hz, H-5), 6.85
bined filtrates (14 mL) were transferred to 50-mL Schlenk (1 H, d, J = 8.1 Hz, H-6), 6.37 (1 H, d, J = 16.1 Hz), 4.19 (2 H,
flasks. Water (1.0 g, 55.0 mmol) and Lipase PS-C “Amano” I dm, J = 24.3 Hz, H-11), 3.90 (3 H, s, H-7), 3.87 (1 H, m, H-12),
(0.100 g) were added to the flasks, and the flasks were shaken 3.58 (2 H, m, H-13). 13C NMR (d6-acetone): δ (ppm) 167.5 (C-
(175 rpm) at 37°C. Aliquots (20 mL) were removed at timed 10), 150.1 (C-4), 148.8 (C-3), 145.8 (C-8), 127.5 (C-1), 123.9
intervals for analytical HPLC and UV spectroscopy (325 nm).
(C-6), 116.1 (C-5), 115.8 (C-9), 111.3 (C-3), 71.0 (C-12), 66.3
Synthesis of feruloylated SBO. The packed-bed bioreactor (C-13), 64.1 (C-11), 56.3 (C-7). Atmospheric pressure CI
synthesis of feruloylated SBO has been detailed previously (APCI)-MS: 266.9 [M – H]– (calcd. C13H16O6: 268.1).
(20). A solution of EF (40 g) dissolved in SBO (160 g) at 60°C
The effluents containing the F2-DAG collected during the
was circulated over a bed of Novozym 435 (34 g) at 60°C for preparative HPLC separations were combined and the solvent
144 h. The reaction progress was monitored by analytical removed under vacuum at 40°C to yield a white solid. The iso-
HPLC (see below). By-products and unreacted starting mater- lated F2-DAG was analyzed by analytical HPLC and was found
ial were removed from the reaction mixtures by molecular dis- to be pure. Yield: 2.3 g. 1H NMR (d6-acetone): δ (ppm) 7.62 (2
tillation at 120°C.
H, d, J = 16.1 Hz, H-8), 7.30 (2 H, s, H-2), 7.10 (2 H, d, J = 6.9
Deacylation of feruloylated SBO. Feruloylated SBO (240 g) Hz, H-5), 6.85 (2 H, d, J = 6.5 Hz, H-6), 6.40 (2 H, d, J = 16.1
and water (14.5 g) were dissolved in 250 mL of 2-methyl-2- Hz), 4.27 (4 H, m, H-11), 4.18 (1 H, m, H-12), 3.88 (3 H, s, H-
propanol in a 1-L Fernbach flask. Lipase PS-C “Amano” I (2.5 7). 13C NMR (d6-acetone): δ (ppm) 167.4 (C-10), 150.1 (C-4),
g) was added and the mixture was shaken (125 rpm) at 37°C 148.8 (C-3), 146.1 (C-8), 127.4 (C-1), 124.0 (C-6), 116.1 (C-
for 144 h. Aliquots (100 mL) were removed at timed intervals 5), 115.6 (C-9), 111.4 (C-3), 68.35 (C-12), 66.0 (C-11), 56.4
for analytical HPLC and UV spectroscopy (325 nm). The sol- (C-7). APCI-MS: 464.9 [M – H + Na], 443.0 [M – H]– (calcd.
vent was removed in vacuo at 60°C for 24 h, and the resultant C23H24O9: 444.14).
oily residue was dissolved in 100 mL of acetone containing
Analytical HPLC. Analyses were performed using a
1.0% acetic acid. The feruloylated monoacylglycerol (F-MAG) Thermo Separation Products (San Jose, CA) HPLC system
and diferuloylated diacylglycerol (F2-DAG) were isolated by consisting of a Spectra System AS3000 autosampler, a Spectra
preparative HPLC (see below). The effluents containing the F- System P4000 pump, a Spectra System UV6000LP detector,
MAG collected during the preparative HPLC separations were an Alltech (Deerfield, IL) 500 ELSD, and a Luna Phenyl-Hexyl
combined and the solvent removed under vacuum at 40°C to column (5 µm, 250 × 4.6 mm; Phenomenex, Torrance, CA).
yield a cloudy, colorless viscous oil. The isolated F-MAG was Solvents were filtered using Whatman 0.45 mm nylon mem-
JAOCS, Vol. 83, no. 9 (2006)

FERULOYLATED ACYLGLYCEROLS FROM VEGETABLE OILS
755
brane filters (Sigma-Aldrich) and degassed using a Thermo feruloylated TAG (F-TAG), feruloylated DAG (F-DAG), and
Separation Products SCM 1000 Membrane Degasser. The fer- feruloylated MAG (F-MAG) as well as trace amounts of fer-
uloylated lipid species were determined using a three-solvent ulic acid (FA), which resulted from the hydrolysis of EF (Fig.
gradient. Solvent A was water (268 mL), methanol (70 mL), 1- 1). Herein, an improved, reversed-phase HPLC–MS method is
butanol (11 mL), and glacial acetic acid (1 mL). Solvent B was discussed that reveals that the lipase-catalyzed transesterifica-
water (93 mL), methanol (245 mL), 1-butanol (33 mL), and tions also produce diferuloylated TAG (F2-TAG) and diferu-
glacial acetic acid (1 mL). Solvent C was methanol. The col- loylated DAG (F2-DAG), which had not been previously iden-
umn was developed at 1.0 mL/min with a 5 min isocratic flow tified.
of 3:1 A/B, a 2-min linear gradient to 100% B, a 5-min isocratic
Recent collaborative interests in the use of F-MAG as a
flow of 100% B, a 2-min linear gradient to 100% C, a 13-min carbon source and selection substrate for the directed evolu-
isocratic flow of 100% C, followed by a 3-min linear gradient tion of feruloyl esterases led to efforts to deacylate the F-
to 3:1 A/B. The UV detector response (325 nm, 7 nm bandpass) DAG and F-TAG mixtures to form F-MAG. The feruloylated
was calibrated with EF. Injection volumes were 10 µL.
acylglycerols were enzymatically hydrolyzed with an excess
HPLC–MS. HPLC–APCI–MS was conducted with a Finni- of water in 2-methyl-2-propanol solutions at 37°C using a sn-
gan-Thermoquest (San Jose, CA) LCQ LC–MS system 1,3-specific enzyme, Lipase PS-C “Amano” I (Fig. 2), which
(AS3000 autoinjector, P4000 HPLC pump, UV6000 PDA de- showed no transesterification or hydrolysis activity towards
tector, LCQ ion-trap mass spectrometer, and a nitrogen genera- ethyl ferulate (19). Fatty acid hydrolysis was facile and
tor) all running under the Xcaliber 1.3 software system. The reached equilibrium after 48 h. Unexpectedly, the deacylation
MS was run with the electrospray ionization (ESI) interface op- of the feruloylated acylglycerols resulted in two feruloylated
erating in the negative ion mode. The source inlet temperature glycerol species, F-MAG and F2-DAG; the latter was incor-
was set at 220°C and the sheath gas rate was set at 90 arbitrary rectly identified as F-MAG using our original water/methanol
units. The MS was optimized by using the autotune feature of
the software while infusing a solution of F2-DAG in with the
effluent of the column and tuning on an atomic mass unit of
531 [M – H]–. The software package was set to collect mass
data between 100 and 1500 amu. Generally the most signifi-
cant sample ions generated under these conditions were [M –
1]–. The LC conditions were the same as those described above
for the analytical system, with the liquid flow exiting the diode
array detector split 1:10 and the low-flow split being directed
into the ESI–MS interface.
Preparative HPLC. Preparative HPLC was used to isolate
F-MAG and F2-DAG. A Shimadzu (Columbia, MD) prepara-
tive HPLC system was used with dual 8A pumps, SIL 10vp au-
toinjector, SPD M10Avp photodiode array detector, and a SCL
10Avp system controller, all operating under the Shimadzu
Class VP operating system. Sample aliquots (2 mL) in 1% ace-
tic acid in acetone (vol/vol) were injected on a Waters (Milford,
MA) Bondapak C18 PrepPak column (15–20 mm, 125 Å, 47 ×
300 mm) in a radial compression module. The column was pre-
equilibrated with 1% acetic acid, 10% acetonitrile, and 89%
acetone at a flow rate of 50 mL/min, and the effluent was mon-
itored at 360 nm. The column was developed to 23% acetoni-
trile over 10 min. Peaks were collected by hand. The procedure
was repeated 40 times to obtain sufficient quantities of purified
F-MAG and F2-DAG.
RESULTS AND DISCUSSION
Synthesis of feruloyl and diferuloyl acylglycerols. SBO was
biocatalytically transesterified with EF to form UV-absorbing
oils (19,20). The Novozym 435-catalyzed transesterifications
were typically conducted by passing solutions of EF in SBO
over a bed of Novozym 435 at 60°C for 144 h. It was previ-
ously reported that the transesterifications resulted in an equi-
librium in which 65% of the EF was converted to a mixture of
FIG. 2. The lipase-catalyzed transesterification of triolein or SBO with EF
followed by lipase-catalyzed hydrolysis. For abbreviations see Figure 1.
JAOCS, Vol. 83, no. 9 (2006)

756
D.L. COMPTON ET AL.
HPLC–MS methods had not revealed the presence of any difer-
uloylated acylglycerol species (19). Therefore, a new
HPLC–MS method was developed to better separate and iden-
tify the feruloylated and diferuloylated acylglycerol species.
HPLC–MS analysis of feruloyl and diferuloyl acylglycerols.
The transesterification of SBO with EF resulted in a mixture of
feruloylated and diferuloylated acylglycerol species. A re-
versed-phase HPLC method using a water/methanol/1-butanol
solvent gradient developed on a phenyl-hexyl column was in-
vestigated for the separation of the feruloylated and diferuloy-
lated acylglycerols. As shown in Figure 3B, the HPLC method
was effective in separating the F-MAG and F2-DAG from the
FA and residual EF. These species were chromatographically
identified using FA and EF standards and the F-MAG and F2-
DAG isolated and purified by preparative HPLC (discussed
earlier). The F-DAG through F-TAG regions [retention time
(Rt) = 18 to 27 min], however, were convoluted with peaks due
to the multiple fatty acids in SBO, making the assignment of
the diferuloylated species ambiguous.
To deconvolute the HPLC–MS chromatogram and simplify
the analysis of the feruloylated species, the lipase-catalyzed
transesterification of triolein with EF was investigated. During
the study of this reaction, it was observed that using glycerol as
a co-reactant enhanced production of the feruloylated species.
Thus, the transesterification of triolein with EF detailed herein
was performed in the presence of glycerol. The chromatogram
of the reaction of a 2:2:1 ratio of EF/triolein/glycerol after
reaching equilibrium (120 h) is shown in Figure 3A. The F-
MAG, F2-DAG, EF, and FA were identified by using standards
as described above. The remaining peaks were identified by
HPLC–APCI–MS in the negative ion mode.
Table 1 lists the peaks obtain by HPLC–MS chromatogra-
phy and their corresponding Rt and major ions. The identifica-
tion of the F-MAG and F2-DAG isolated by preparative HPLC
and characterized by 1H and 13C NMR spectroscopy was con-
firmed by HPLC–MS. The major ions of the F2-DAG species
were 23 mu greater than their calculated exact mass, which cor-
responds to the addition of a Na+ atom to the negative F2-DAG
ion. Ion contaminants from salts (e.g., Na+ and Li+) often form
adducts with the major ion (21).
FIG. 3. Analytical reversed-phase HPLC chromatograms of the
Novozym 435-catalyzed transesterification of (A) triolein and (B) SBO
with EF. The chromatograms were obtained using a Luna phenyl-hexyl
column developed at 1.0 mL/min with a gradient of a tertiary solvent
system consisting of water, methanol and n-butanol monitored at 325
nm. See Figure 1 for acronym definitions.
and acetone/acetonitrile, reversed-phase, C8 column, HPLC
methods (19). The isolation of the two feruloylated glycerol
species by preparative HPLC was achieved using an isocratic
development of a C18 column with an amended acetone/ace-
1
tonitrile solvent system. H NMR analysis of the isolated
solids allowed for the unambiguous identification of 1(3)-fer-
uloyl-sn-glycerol (F-MAG) and 1,3-diferuloyl-sn-glycerol
(F2-DAG). The F-MAG spectrum consisted of three unique
sets of glycerol protons in a 2:1:2 ratio (H-11/H-12/H-13), in-
dicating an sn-1(3) substitution of the glycerol backbone. The
ratio of the feruloyl methyl peak (H-7) to the glycerol protons
confirmed a single feruloyl moiety per glycerol. The F2-DAG
spectrum showed two sets of glycerol protons (H-11/H-12) in
a 4:1 ratio, consistent with an sn-1,3 substitution pattern. The
ratio of the feruloyl methyl peak (H-7) to the glycerol protons
confirmed two feruloyl moieties per glycerol. Identification
of the F-MAG and F2-DAG was confirmed by HPLC-MS
(discussed shortly).
The presence of the previously undetected diferuloyl
monooleoylglycerol (F2MOG) species was confirmed by
HPLC–MS using the new reversed-phase pheny-hexyl column
method. The F2MOG eluted as a single peak (Rt = 20.2 min; Fig.
3A) with the expected mass (Table 1) followed by a small shoul-
der. The major ion detected for the shoulder also corresponded
to the mass for F2MOG. Because Novoym 435 has an sn-1(3)-
positional acylation preference, it was assumed that the major
F2MOG peak corresponded to the 1,3-diferuloyl-2-oleoyl-sn-
glycerol isomer and the shoulder was the less abundant 1,2-
diferuloyl-3-oleoyl-sn-glycerol. Similarly, the other diferuloy-
lated species, F2-DAG, eluted as two peaks, each possessing the
same mass. The larger peak (Rt = 11.1 min; Fig. 3A) was identi-
fied by the F2-DAG standard isolated by preparative HPLC and
was confirmed by 1H NMR to be the 1,3-diferuloyl-sn-glycerol
(discussed earlier). Thus, the less abundant peak was assumed
The unexpected discovery of the F2-DAG species indicated
that a significant quantity of the F2-TAG must be produced dur-
ing the transesterification of SBO with EF. However, previous
JAOCS, Vol. 83, no. 9 (2006)

FERULOYLATED ACYLGLYCEROLS FROM VEGETABLE OILS
757
TABLE 1
Feruloylated Species Obtained from the Novozym 435-Catalyzed Transesterification of Triolein
with Ethyl Ferulate (EF)
Speciesa
Rt (min)b
Exact mass
Major ion (m/z)c
Peak area (%)d
F-MAG
FMOG
FDOG
F2-DAG
F2-DAGe
F2MOG
F2MOGe
6.7
18.5
25.3
11.1
11.4
268.1
532.3
796.6
444.1
444.1
708.1
708.1
266.9 [M – H]–
8.2
29.6
11.7
13.8
0.6
531.1 [M – H]–
795.3 [M – H]–
464.9 [M – H + Na]
465.0 [M – H + Na]
714.3 [M – H + Li]
707.1 [M – H]–
20.2
1.1
0.9
20.2 shf
aFeruloylated species obtained after the enzyme-catalyzed reaction of 2:2:1 mole ratio of EF/triolein/glycerol at 60°C (120
h). FA, ferulic acid; F-MAG, feruloylated MAG; FMOG, feruloyl monooleoylglycerol; FDOG, feruloyl dioleoylglycerol; F2-
DAG, diferuloylated DAG; F2MOG, diferuloyl monooleoylglycerol.
bRetention times were determined using the UV6000 detector (325 nm) during HPLC–MS.
cMajor ions detected by HPLC–MS in negative ion mode.
dThe percent peak area of the species related to the total peak area of all feruloylated species, including FA and EF, as de-
tected by HPLC using the UV detector (325 nm).
esn-1,2-Diferuloyl positional isomers.
fShoulder.
to be 1,2-diferuloyl-sn-glycerol. The elution of the 1,3-F2-DAG ratio (Table 2). The FDOG, 1-feruloyl-dioleoyl-sn-glyerol,
slightly before the 1,2-F2DAG is consistent with the reversed- when hydrolyzed by the predominantly sn-1,3-specific B. cepa-
phase HPLC elution sequence of 1,3-dipalmitoyl glycerol be- cia lipase, which showed no hydrolysis activity toward EF
fore 1,2-dipalmitoyl glycerol and 1,3-dilinoleoyl glycerol be- (19), would be expected to yield the 1-feruloyl-2-oleoyl-sn-
fore 1,2-dilinoleoyl glycerol (6). Therefore, it can be concluded glycerol FMOG isomer. The HPLC chromatogram of the hy-
that the sn-1,3-diferuloyl isomers eluted before the sn-1,2-difer- drolyzed reaction mixture after 24 h (data not shown) showed
uloyl isomers, further suggesting that the larger F2MOG peak a large increase in the FMOG shoulder (Rt = 18.5 min) and
was the 1,3-diferuloyl-2-oleoyl-sn-glycerol.
complete consumption of the FDOG. This strongly suggested
Feruloyl monooleoylglycerol (FMOG) species eluted as a that the FMOG shoulder was the 1-feruloyl-2-oleoyl-sn-glyc-
small shoulder preceding a single peak (Rt = 18.5 min; Fig. erol isomer. Thus, the FMOG main peak was designated as the
3A). The main peak possessed a major ion of the expected 1-feruloyl-3-oleoyl-sn-glycerol isomer. The FMOG main peak
mass, but the MS data for the shoulder were inconclusive. It was essentially consumed within 24 h under the hydrolysis
was likely, however, that the shoulder was also an FMOG iso- conditions and would be expected to be converted to F-MAG.
mer. Based on the sn-1(3) positional acylation preference of Indeed, the F-MAG peak increased relative to the consumption
Novozym 435, it was concluded that the feruloyl group of the of the FMOG main peak, further supporting that the FMOG
various monoferuloyl acylglycerol species occupies the sn- main peak was the1-feruloyl-3-oleoyl-sn-glycerol.
1(3)-position. Hydrolysis of the feruloylated oleoylglycerols
The results just presented show that the reversed-phase,
offered evidence as to which FMOG positional isomer, the 1- phenyl-hexyl column, HPLC–MS method allowed for the sep-
feruloyl-3-monooleoyl-sn-glycerol or 1-feruloyl-2-oleoyl-sn- aration and unambiguous identification of the various feruloy-
glycerol, corresponded to the shoulder and which corresponded lated and diferuloylated oleoylglycerols and their positional
to the main peak. Consider the Novozym 435-catalyzed trans- isomers formed during the lipase-catalyzed transesterification
esterification of triolein with EF without glycerol, which pro- of triolein with EF. Although the many feruloylated acylglyc-
duced the highest feruloyl dioleoylglycerol (FDOG) to FMOG erols formed during the transesterification of SBO with EF (Rt
TABLE 2
Ratio of Reactants and the Corresponding Distribution of Feruloylated Mono- and Dioleoyl Glycerol Products
Product species distribution (%)a
Triolein/EF/glycerol
(mole ratio)
F-MAG
(6.7)b
F2-DAG
(11.1)b
F2-DAG
(11.4)b
FMOGc
FMOG
(18.5)b
F2MOG
(19.2)b
F2MOGc
(19.2)b
FDOG
(25.3)b
Trial
(18.5)b
A
B
C
D
1:1:1
2:2:1
4:4:1
1:1:0
18.6
8.2
3.0
0.4
12.2
13.8
15.7
6.2
0.3
0.6
0.6
0.0
4.0
3.9
3.0
1.1
32.5
29.6
21.7
11.0
0.9
1.1
2.3
8.2
0.6
0.9
1.1
0.0
7.5
11.7
16.0
19.7
aThe product species distributions at reaction equilibrium were determined as the percent peak area of the species related to the total peak area of all feru-
loylated species, including FA and EF, as detected by HPLC using the UV detector (325 nm). See Table 1 for acronym definitions.
bThe Rt (min) corresponding to the peaks in Figure 2A.
cRatio of shoulder to the main peak (see Fig. 2A).
JAOCS, Vol. 83, no. 9 (2006)

758
D.L. COMPTON ET AL.
erols by High-Performance Liquid Chromatography with a Sil-
ver Ion Column, J. Chromatogr. A 454:273–284 (1988).
10. Mondello L., P.Q. Tranchida, V. Stanek, P. Jandera, G. Dugo,
and P. Dugo, Silver-ion Reversed-Phase Comprehensive Two-
Dimensional Liquid Chromatography Combined with Mass
Spectrometric Detection in Lipidic Food Analysis, Ibid.
1098:91–98 (2005).
11. Hopia, A.I., V.I. Piironen, P.E. Koivistoinen, and L.E.T.
Hyvoven, Analysis of Lipid Classes by Solid-Phase Extraction
and High-Performance Size-Exclusion Chromatography, J. Am.
Oil Chem. Soc. 69:772–776 (1992).
= 18.0–28.0 min, Fig. 3B) could not be as cleanly separated
using the phenyl-hexyl HPLC–MS column method because of
SBO’s multiple fatty acids, the results from the triolein esterifi-
cations were used to assign the peaks to classes of feruloylated
SBO glycerols. Figure 3B shows that the F-DAG, F2-TAG, and
F-TAG eluted at times similar to those of their corresponding
feruloylated oleoylglycerols. This allowed for quantification of
the three groups of feruloylated acylglycerols, which was not
possible using previous HPLC methods (19,20).
12. Lin, J.-T., and T.A. McKeon (eds.), HPLC of Acyl Lipids, HNB
Publishing, New York, 2005, 576 pp.
13. Marcato, B. and G. Cecchin, Analysis of Mixtures Containing
Free Fatty Acids and Mono-, Di- and Triglycerides by High-Per-
formance Liquid Chromatography Coupled with Evaporative
Light-Scattering Detection, J. Chromatogr. A. 730:83–90
(1996).
ACKNOWLEDGMENTS
The authors thank Kendra Brandon, Leslie Smith, Ray Halloway,
and Sandy Duval for their excellent technical assistance.
14. Foglia, T.A., and K.C. Jones, Quantitation of Neutral Lipid Mix-
tures Using High-Performance Liquid Chromatography with
Light Scattering Detection, J. Liq. Chromatogr. Relat. Technol.
20:1829–1838 (1997).
15. Berner, D., and A. Dieffenbacher, Determination of Mono- and
Diacylglycerols in Edible Oils and Fats by High-Performance
Liquid Chromatography and Evaporative Light Scattering De-
tector: Results of Collaborative Studies and the Standardized
Methods, Pure Appl. Chem. 71:1983–1991 (1999).
16. Knothe, G., Analytical Methods Used in the Production and Fuel
Quality Assessment of Biodiesel, Trans. ASAE 44:193–200
(2001).
17. Mullin, W.J., and J.P.H. Emery, Determination of Alkylresorci-
nols in Cereal-Based Foods, J. Agric. Food. Chem. 40:2127–2130
(1992).
18. Institute for Biochemistry and Molecular Biology, University of
Wroclaw, Poland, Resorcinolic Lipids, http://www.ibmb.uni.
wroc.pl/liprez.htm (accessed March 2006).
19. Compton, D.L., J.A. Laszlo, and M.A. Berhow, Lipase-Cat-
alyzed Synthesis of Ferulate Esters, J. Am. Oil. Chem. Soc.
77:513–519 (2000).
20. Laszlo, J.A., D.L. Compton, F.J. Eller, S.L. Taylor, and T.A. Is-
bell, Packed-Bed Bioreactor Synthesis of Feruloylated Monoa-
cyl- and Diacylglycerols: Clean Production of a “Green” Sun-
screen, Green Chem. 5:382–386 (2003).
21. Takayama, M., T. Fukai, T. Nomura, and T. Yamauchi, Forma-
tion and Fragmentation of the [M + Na]+ Ion of Glycosides in
Fast Atom Bombardment Mass Spectrometry, Org. Mass Spec-
trom. 26:655–659 (1991).
REFERENCES
1. Hills, G., Industrial Use of Lipases to Produce Fatty Acid Es-
ters, Eur. J. Lipid Sci. Technol. 105:601–607 (2003).
2. Knothe, G., R.O. Dunn, and M.O. Bagby, Biodiesel: The Use of
Vegetable Oils and Their Derivatives as Alternative Diesel
Fuels, in Fuels and Chemicals from Biomass, edited by B.C.
Saha and J. Woodward, American Chemical Society, Washing-
ton, DC, 1997, pp. 172–208.
3. Dunn, R.O., G. Knothe, and M.O. Bagby, Recent Advances in
the Development of Alternative Diesel Fuel from Vegetable Oil
and Animal Fat, Recent Res. Dev. Oil Chem. 1:31–56 (1997).
4. Erhan, S. (ed.), Industrial Uses of Vegetable Oils, AOCS Press,
Champaign, IL, 2005.
5. Liu, J., T. Lee, E. Bobik, M. Guzmanharty, and C. Hastilow,
Quantitative Determination of Monoglycerides and Diglycerides
by High Performance Liquid Chromatography and Evaporative
Light Scattering Detection, J. Am. Oil Chem. Soc. 70:343–347
(1993).
6. Perona, J.S., and V. Ruiz-Gutierrez, Simultaneous Determina-
tion of Molecular Species of Monoacylglycerols, Diacylglyc-
erols, and Triacylglycerols in Human Very-Low-Density
Lipoproteins by Reversed-Phase Liquid Chromatography, J.
Chromatogr. B 785:89–99 (2003).
7. Bergqvist, M.H.J., and P. Kaufmann, A Multivariate Optimiza-
tion of Triacyglycerol Analysis by High-Performance Liquid
Chromatography, Lipids 28:667–675 (1993).
8. Nikolova-Damyanova, B., Lipid Analysis by Silver Ion Chro-
matography, in Advances in Lipid Methodology—Five, edited
by R.O. Adlof, Oily Press, Bridgewater, 2003, pp. 43–123.
9. Christie W.W., Separation of Molecular Species of Triacylglyc-
[Received December 20, 2005; accepted June 25, 2006]
JAOCS, Vol. 83, no. 9 (2006)