Synthesis of steryl ferulates with various sterol structures and comparison of their antioxidant activity

Article abstract of DOI:10.1016/j.foodchem.2014.07.119

Steryl ferulates synthesised from commercial sterols as well as commercial oryzanol were used to better understand how structural features affect antioxidant activity in vitro by the ABTS⁺radical decolorization assay, by oxidative stability ind

Full text of DOI:10.1016/j.foodchem.2014.07.119

Food Chemistry 169 (2015) 92–101
Contents lists available at ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Synthesis of steryl ferulates with various sterol structures
and comparison of their antioxidant activity
Jill K. Winkler-Moser a, , Hong-Sik Hwang , Erica L. Bakota , Debra A. Palmquist
⇑
a
a
b
a
United States Department of Agriculture,1 Agricultural Research Service, National Center for Agricultural Utilization Research, Functional Foods Research,
815 N University Street, Peoria, IL 61604, United States
United States Department of Agriculture, Agricultural Research Service, Midwest Area, 1815 N University Street, Peoria, IL 61604, United States
1
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Steryl ferulates synthesised from commercial sterols as well as commercial oryzanol were used to better
Received 21 October 2013
Received in revised form 18 April 2014
Accepted 25 July 2014
understand how structural features affect antioxidant activity in vitro by the ABTSꢀ+ radical decolorization
assay, by oxidative stability index (OSI) of soybean oil, and by analysis of antioxidant activity during fry-
ꢀ+
ing. Steryl ferulates inhibited the ABTS radical by 6.5–56.6%, depending on their concentration, but were
Available online 4 August 2014
less effective, especially at lower concentrations, than ferulic acid. Ferulic acid and steryl ferulates had
either no effect, or lowered the OSI of soybean oil by up to 25%, depending on the concentration. In their
evaluation as frying oil antioxidants, steryl ferulates with a saturated sterol group had the best antioxi-
dant activity, followed by sterols with one double bond in the C5 position. The results indicate that a
dimethyl group at C4 as well as a C9,C19 cyclopropane group, as found in oryzanol, negatively affects
antioxidant activity in frying oils.
Keywords:
Steryl ferulates
Phytosterols
Soybean oil
Frying
Antioxidant activity
Antipolymerisation activity
Published by Elsevier Ltd.
1
. Introduction
Steryl ferulates are composed of ferulic acid esterified to plant
2007; Xu & Godber, 2001). The hydrophobic sterol group allows
steryl ferulates to be soluble in oils, making them excellent candi-
dates for antioxidants in oil-rich foods. In addition, because of the
increased molecular weight of steryl ferulates compared to ferulic
acid, steryl ferulates are less likely to evaporate at high tempera-
tures such as in frying oils and in other high temperature food
processes. Steryl ferulates are also less susceptible to thermal
decomposition as compared to non-esterified ferulic acid
(Shopova & Milkova, 2000). Nyström et al. (2007) demonstrated
that 0.5% and 1.0% sitostanyl ferulate inhibited polymer formation
in stripped high oleic sunflower oil (HOSUN) heated to 180 °C by
27% and 43%, respectively. In previous frying studies, we found that
corn steryl ferulates inhibited polymerisation of soybean oil by up
to 70%, whereas oryzanol was much less effective (Winkler-Moser,
Rennick, Palmquist, Berhow, & Vaughn, 2012). Corn steryl ferulates
also protected the endogenous tocopherols in soybean oil during
frying, whereas oryzanol did not. It was also noted that oryzanol
degraded at a faster rate during frying compared to corn steryl fer-
ulates. It was hypothesised that higher stability was the reason for
the better anti-polymerisation activity; however, in a heating
study (180 °C, in soybean oil), corn steryl ferulates actually
degraded at a faster rate compared to oryzanol, but still had better
antipolymerisation activity (Winkler-Moser, Rennick, Hwang,
Berhow, & Vaughn, 2013), indicating that the differences in activity
may not be completely attributed to their stability. In fact, Nyström
et al. (2005) observed stronger antioxidant activity for wheat and
sterols and found in the inner pericarp and aleurone layer of cere-
als grains such as corn, wheat, rice, rye, and triticale (Moreau,
Singh, Nuñez, & Hicks, 2000; Norton, 1995; Seitz, 1989). Steryl
ferulates from rice bran oil, which are called
c
-oryzanol (oryzanol
0
for short), are mainly composed of two 4,4 -dimethylsterols, cyclo-
artenol and 24-methylene-cycloartanol, as well as the 4-desmeth-
ylsterol, campesterol, and low amounts of other 4-desmethyl
sterols (Xu & Godber, 1999). Corn steryl ferulates consist mainly
of two saturated 4-desmethylsterols, sitostanol, and campestanol,
with lower amounts of other 4-desmethylsterols such as sitosterol
and campesterol (Seitz, 1989).
Ferulic acid is a well-known phenolic antioxidant, thus steryl
ferulates have antioxidant activity due to the radical scavenging
activity of the phenolic component (Juliano, Cossu, Alamanni, &
Piu, 2005; Kikuzaki, Hisamoto, Hirose, Akiyama, & Taniguchi,
2
002; Nyström, Achrenius, Lampi, Moreau, & Piironen, 2007;
Nyström, Mäkinen, Lampi, & Piironen, 2005; Suh, Yoo, & Lee,
⇑
Corresponding author. Tel.: +1 309 681 6390; fax: +1 309 681 6685.
E-mail address: Jill.Moser@ars.usda.gov (J.K. Winkler-Moser).
Mention of trade names or commercial products in this publication is solely for
1
the purpose of providing specific information and does not imply recommendation or
endorsement by the U.S. Department of Agriculture. USDA is an equal opportunity
provider and employer.
http://dx.doi.org/10.1016/j.foodchem.2014.07.119
0
308-8146/Published by Elsevier Ltd.

J.K. Winkler-Moser et al. / Food Chemistry 169 (2015) 92–101
93
rye steryl ferulates compared to oryzanol and cycloartenyl ferulate
in methyl linoleate oxidised at 40 °C, which is not likely to be
related to thermal stability.
and brassicasterol are the four most common plant sterols, and
they all have a C5 double bond. Since the majority of corn steryl
ferulates have saturated (no double bonds) phytosterols, b-choles-
tanol (B) and sitostanol (C, also called stigmastanol) were chosen
as model compounds for this group. Finally, lanosterol (E) was cho-
sen to study the effect of the dimethyl group at C4, which is a
structural feature that is also found in the two major steryl feru-
lates found in oryzanol, cycloartenol (F) and 24-methylene cyclo-
artanol (G). These two sterols were not commercially available in
the quantities needed for this study, so commercially available
oryzanol was studied instead. Also shown in Scheme 1 is campes-
terol (H), which together with sitosterol makes up a low percent-
age of corn steryl ferulates (ꢀ14%), but makes up about 31% of
oryzanol.
Since the only difference between oryzanol and corn steryl fer-
ulates are in the structures of the phytosterol moieties, the objec-
tive of this research was to conduct studies to further understand
the effect of structural features of the phytosterol head group on
antioxidant activity. Since isolation and purification of the quanti-
ties needed to carry out these analyses would require large
amounts of starting material, as well as solvents, we chose to syn-
thesise steryl ferulates using pure phytosterols. Previously
published synthesis techniques (Condo, Baker, Moreau, & Hicks,
2
001; Ebenezer, 1991) were improved for time, solvent consump-
tion, and yield, and the antioxidant activity of synthetic steryl
ferulates as well as oryzanol and ferulic acid were evaluated
in vitro and by analysis of oxidative stability index in soybean oil,
as well as in frying studies in soybean oil.
2. Materials and methods
The structures of the steryl ferulates that were synthesised are
shown in Scheme 1. Cholesterol (A) and sitosterol (D) were chosen
to represent sterols with a double bond at the C5-position, which
should provide information about the effect of this double bond
on steryl ferulate activity. Sitosterol, stigmasterol, campesterol,
2.1. Materials
8,24,(5-a)-Cholestadien-4,4,14-a-trimethyl-3-b-ol (lanosterol,
ꢀ65%) was purchased from Steraloids Inc. (Newport, RI). ABTS
Scheme 1. Structures of synthetic steryl ferulates: (A) cholesterol, (B) cholestanol, (C) sitostanol, (D) sitosterol, (E) lanosterol, and structures of steryl ferulates in oryzanol: (F)
cycloartenol, (G) 24-methylene cycloartanol, (H) campesterol. R = trans-ferulic acid.

9
4
J.K. Winkler-Moser et al. / Food Chemistry 169 (2015) 92–101
0
Ò
(
(
2,2 -azinobis-(3-ethylbenzothiazoline-6-sulphonic acid)), trolox
3
NMR (CDCl ) 11.9, 18.7, 19.3, 20.6, 21.1, 22.6, 22.8, 23.8, 24.3,
6-hydroxy-2,5,7,8-tetramethlchroman-2-carboxylic acid), trans-
27.9, 28.0, 28.2, 31.9, 31.9, 35.8, 36.2, 36.6, 37.0, 38.2, 39.5, 39.8,
42.3, 50.1, 52.9, 56.2, 56.7, 74.2, 111.2, 118.9, 121.2, 122.8, 123.2,
133.5, 139.6, 141.4, 143.7, 151.4, 166.2, 168.7.
ferulic acid, pyridine, cholesterol (94%), cholestanol (P95%),
b-sitosterol (from soybean, >70%), stigmastanol (>95%), 4-(dimeth-
ylamino)pyridine (DMAP), acetic anhydride, dicyclohexylcarbodi-
imide (DCC), and potassium carbonate were purchased from
Sigma–Aldrich (St. Louis, Missouri, USA). Methanol, ethyl acetate,
dichloromethane, chloroform, hexanes, and hydrochloric acid were
purchased from Fisher Scientific (Fair Lawn, NJ, USA). Oryzanol
98%) was from Spectrum (Gardena, CA, USA). Silica gel 60 (70–
30 mesh) used for column chromatography was purchased from
EMD Chemical Inc. (Darmstadt, Germany). Thin layer chromatogra-
phy (TLC) plates (MK6F silica gel 60 Å) were purchased from What-
man (Clifton, NJ, USA). Refined, bleached, deodorized soybean oil
was kindly provided by Archer Daniels Midland (Decatur, IL, USA).
For synthesis of 3-O-(trans-4-O-acetylferuloyl)-b-sitosterol, the
same procedure used for 3-O-(trans-4-O-acetylferuloyl)-b-choles-
tanol provided 7.3 g (96%) of the product after column chromatog-
raphy and 6.9 g (90%) after recrystallization from methanol/
1
13
dichloromethane as a white solid. H and C NMR data were con-
sistent with published literature values (Condo et al., 2001): mp
(
2
ꢁ1
171–172 °C [lit. 162 °C (Condo et al., 2001)]; IR (cm ) 2935,
2867, 1760, 1710, 1637, 1259, 1172, 1154, 1031, 1009; 1H NMR
3
(CDCl ) d 0.70 (s, 3H), 0.83 (d, 3H, J = 6.9), 0.85–0.89 (m, 6H),
0.95 (d, 3H, J = 6.6), 0.97–1.05 (m, 3H), 1.09–2.09 (m, 28H), 2.33
(s, 3H), 2.42 (m, 1H), 3.87 (s, 3H), 4.76 (m, 1H), 5.42 (br d, 1H),
6
.38 (d, J = 16.0 Hz, 1H), 7.05 (d, J = 8.6 Hz, 1H), 7.12 (m, 2H),
1
3
2
2
1
.2. Methods
7.64 (d, J = 15.9 Hz, 1H); C NMR (CDCl
3
) d 11.9, 12.0, 18.8, 19.1,
1
9.3, 19.8, 20.6, 21.1, 23.1, 24.3, 26.1, 27.9, 28.3, 29.2, 31.9, 31.9,
.2.1. Synthesis of acetylferulic sterols
Acetic anhydride (18.7 mL, 0.20 mol) was added dropwise over
0 min to the stirred solution of trans-ferulic acid (35.0 g, 0.18 mol)
34.0, 36.2, 36.6, 37.0, 38.2, 39.7, 42.3, 45.9, 50.1, 55.9, 56.1, 56.7,
74.2, 111.2, 118.9, 121.2, 122.8, 123.2, 133.5, 139.6, 141.4, 143.7,
151.4, 166.2, 168.7.
and DMAP (0.14 g, 0.0011 mol) in dry pyridine (109 mL) at 0 °C
under nitrogen. The mixture was stirred at 0 °C for 10 min, warmed
to room temperature slowly and stirred for 4 h at room tempera-
ture. A 1:1 mixture of ice and water (100 g) was added and the
mixture was stirred for 10 min maintaining the temperature at
For synthesis of 3-O-(trans-4-O-acetylferuloyl)-8,24,(5-
a)-
cholestandien-4,4,14- -trimethyl-3-b-ol, the same procedure
a
used for 3-O-(trans-4-O-acetylferuloyl)-b-cholestanol provided
6.4 g (85%) of the product after column chromatography and
5.8 g (77%) after recrystallization from methanol/dichlorometh-
ꢁ
1
0
°C and neutralized by slowly adding 2 N HCl. The precipitate
was filtered and washed with de-ionised water. Methanol
210 mL) was added to the resulting white solid and the mixture
ane as a white solid. mp 154–165 °C; IR (cm ) 2939, 2866,
1
3
11758, 1699, 1634, 1286, 1154, 1123, 1031; H NMR (CDCl ) d
(
0.72 (s, 3H), 0.85–0.89 (m, 3H), 0.90 (s, 3H), 0.91–0.94 (m, 3H),
0.95 (s, 3H), 0.99 (s, 3H), 1.05 (s, 3H), (m, 12H), 1.62 (s, 3H),
1.70 (s, 3H), 1.72–2.09 (m, 11H), 2.33 (s, 3H), 2.42 (m, 1H),
3.88 (s, 3H), 4.67 (m, 1H), 5.12 (m, 1H), 6.40 (d, J = 16.0 Hz,
1H), 7.06 (d, J = 8.0 Hz, 1H), 7.12 (m, 2H), 7.63 (d, J = 15.9 Hz,
was stirred at 55–60 °C for 30 min and then at room temperature
for 1 h. The solid was filtered and dried to give the product as white
1
13
solid (39.0 g, 91.6%). H and C NMR data were consistent with
published literature values (Ebenezer, 1991).
1
3
For synthesis of 3-O-(trans-4-O-acetylferuloyl)-b-cholestanol,
to a mixture of b-cholestanol (5.0 g, 12.9 mmol) and trans-4-O-ace-
tylferulic acid (4.6 g, 19.5 mmol) in dry dichloromethane (100 mL),
DMAP (0.5 g, 4.1 mmol) and DCC (4.1 g, 19.9 mmol) were subse-
quently added at 0 °C under nitrogen. The mixture was stirred at
3
1H); C NMR (CDCl ) d 15.8, 16.7, 17.6, 18.2, 18.6, 19.2, 20.6,
21.0, 22.6, 22.8, 24.1, 24.3, 24.3, 24.9, 25.7, 26.4, 28.0, 28.2,
30.8, 31.0, 35.3, 36.3, 36.4, 36.9, 38.1, 44.5, 81.1, 111.2, 119.1,
121.2, 123.2, 125.3, 133.5, 134.3, 141.3, 143.5, 151.4, 166.6,
168.7.
0
°C for 30 min and then at room temperature for 1 h. Hexanes
For synthesis of 3-O-(trans-4-O-acetylferuloyl)-stigmastanol,
the same procedure used for 3-O-(trans-4-O-acetylferuloyl)-b-cho-
lestanol provided 6.6 g (86%) of the product after column chroma-
tography and 6.4 g (84%) after recrystallization from methanol/
(
(
300 mL) was added and the insoluble solid (1,3-dicyclohexylurea
DCU)) was removed by filtration. Solvent was evaporated and the
residue was subjected to column chromatography on silica gel
with dichloromethane/hexanes/ethyl acetate (4:16:1) to give the
product (7.1 g, 91%). Further purification by recrystallization from
methanol/dichloromethane provided 6.6 g (84%) of the product
as a white solid. mp 147–149 °C; IR (cm ) 2931, 2849, 1758,
1
ꢁ1
dichloromethane as a white solid. mp 163–164 °C; IR (cm
)
2940, 2866, 1754, 1710, 1638, 1263, 1224, 1171, 1154, 1033,
1
3
1010; H NMR (CDCl ) d 0.67 (s, 3H), 0.74–2.06 (m, 47H), 2.33 (s,
ꢁ1
3H), 3.87 (s, 3H), 4.84 (m, 1H), 6.37 (d, J = 16.0 Hz, 1H), 7.05 (d,
698, 1261, 1178, 1155, 1125, 1031, 1013; 1H NMR (CDCl
.67 (s, 3H), 0.86 (s, 3H), 0.88 (d, 3H, J = 2.2 Hz), 0.89 (d, 3H,
) d
J = 8.7 Hz, 1H), 7.11 (m, 2H), 7.63 (d, J = 15.9 Hz, 1H); C NMR
13
3
0
3
(CDCl ) d 12.0, 12.1, 12.3, 18.7, 19.0, 19.8, 20.6, 21.2, 23.1, 24.2,
J = 2.2 Hz), 0.92 (d, 3H, J = 6.7 Hz), 0.96–2.04 (m, 31H), 2.32 (s,
H), 3.86 (s, 3H), 4.85 (m, 1H), 6.37 (d, J = 16.0 Hz, 1H), 7.05 (d,
26.1, 27.6, 28.3, 28.6, 29.2, 32.0, 33.9, 34.1, 35.5, 36.2, 36.8, 40.0,
42.6, 44.7, 45.9, 54.3, 55.9, 56.2, 56.4, 74.0, 111.2, 119.1, 121.2,
123.2, 133.5, 143.5, 151.4, 166.3, 168.7.
3
13
J = 8.7 Hz, 1H), 7.12 (m, 2H), 7.63 (d, J = 16.0 Hz, 1H); C NMR
CDCl ) d 12.1, 12.3, 18.7, 20.6, 21.2, 22.6, 22.8, 23.9, 24.2, 27.6,
8.0, 28.2, 28.6, 32.0, 34.1, 35.5, 35.8, 36.2, 36.8, 39.5, 40.0, 42.6,
(
3
2
4
1
2.2.2. Synthesis of steryl ferulates
4.7, 53.4, 54.3, 55.9, 56.3, 56.4, 74.0, 111.2, 119.1, 121.2, 123.2,
33.5, 141.3, 143.5, 151.4, 166.3, 168.7.
For synthesis of 3-O-(trans-4-feruloyl)-b-cholestanol (cholesta-
nyl ferulate), K
of 3-O-(trans-4-O-acetylferuloyl)-b-cholestanol (5.45 g, 9.0 mmol)
in CHCl -MeOH (2:1, 150 mL) and the mixture was refluxed for
6 h. The reaction mixture was cooled down to room temperature
and saturated aqueous NH Cl (15 mL) was added. Dichlorometh-
2 3
CO (0.24 g, 1.74 mmol) was added to the solution
For synthesis of 3-O-(trans-4-O-acetylferuloyl)-cholesterol, the
same procedure used for 3-O-(trans-4-O-acetylferuloyl)-b-choles-
tanol provided 7.1 g (91%) of the product after column chromatog-
raphy and 6.5 g (83%) after recrystallization from methanol/
3
4
ꢁ1
dichloromethane as a white solid. mp 167–168 °C; IR (cm
)
ane (200 mL) and water (200 mL) were added and the organic layer
was separated, washed with water (200 mL) and then with brine
(100 mL), dried over MgSO , and filtered. Short column chromatog-
4
2
1
934, 2867, 1759, 1701, 1635, 1262, 1175, 1155, 1124, 1031,
1
006; H NMR (CDCl
3
) d 0.70 (s, 3H), 0.88 (d, J = 2.2 Hz, 3H), 0.89
(
1
(
7
d, J = 2.2 Hz, 3H), 0.94 (d, J = 6.6 Hz, 3H), 0.95–1.06 (m, 3H),
.07(s, 3H), 1.08–2.09 (m, 24H), 2.33 (s, 3H), 2.42 (m, 1H), 3.87
s, 3H), 4.76 (m, 1H), 5.42 (br m, 1H), 6.38 (d, J = 15.9 Hz, 1H),
raphy (50 g silica gel) with 1:3:1 of dichloromethane/hexanes/
ethyl acetate followed by recrystallization from dichlorometh-
ane-methanol gave a white solid (4.91 g, 97%). mp 132–133 °C;
1
3
ꢁ1
.06 (d, J = 8.6 Hz, 1H), 7.12 (m, 2H), 7.63 (d, J = 15.9 Hz, 1H);
C
IR (cm ) 3402, 2931, 2866, 2849, 1706, 1636, 1263, 1172, 1154,

J.K. Winkler-Moser et al. / Food Chemistry 169 (2015) 92–101
95
1
032; 1H NMR (CDCl
J = 2.2 Hz), 0.89 (d, 3H, J = 2.2 Hz), 0.92 (d, 3H, J = 6.7 Hz), 0.95–
.04 (m, 38H), 2.32 (s, 3H), 3.92 (s, 3H), 4.83 (m, 1H), 5.95 (s,
3
) d 0.67 (s, 3H), 0.86 (s, 3H), (0.89 (d, 3H,
2.2.3. Analysis of synthesised products
Melting point was measured with OptoMelt Automated melting
point system (Stanford Research systems, Sunnyvale, CA, USA). IR
spectrum was obtained with Nexus 470 FT-IR equipped with Smart
Orbit (Thermo Electron Corporation, Madison, WI, USA). FT-IR data
were analysed with Monic 8.2 software (Thermo Fisher Scientific
2
1
H), 6.28 (d, J = 15.9 Hz, 1H), 6.92 (d, J = 8.2 Hz, 1H), 7.4 (d,
1
3
J = 1.7 Hz, 1H), 7.07 (m, 1H), 7.60 (d, J = 15.9 Hz, 1H); C NMR
CDCl ) d 12.1, 12.3, 18.7, 21.2, 22.6, 22.8, 23.9, 24.2, 27.6, 28.0,
8.3, 28.7, 32.0, 34.2, 35.5, 35.8, 36.2, 36.8, 39.5, 40.0, 42.6, 44.7,
(
3
1
2
5
1
Inc., Waltham, MA, USA). H NMR spectra were acquired on a Bru-
4.3, 55.9, 56.3, 56.4, 73.7, 109.3, 114.7, 116.2, 123.0, 127.1,
44.4, 146.8, 147.9, 166.8. Analysis by TLC showed one spot.
For synthesis of 3-O-(trans-4-feruloyl)-cholesterol (cholesteryl
ker (Billerica, MA, USA) Avance 500 spectrometer operating at
1
3
500 MHz. C NMR experiment was also done at 125 MHz. Spin-
Works 3.1.7 (Marat, 2011) software was used for analysis of
NMR signals. Purity was determined by TLC (silica gel, dichloro-
methane/hexanes/ethyl acetate 4:16:1 v/v/v). Since commercial
phytosterol standards often contain impurities, and because the
commercial sitosterol and lanosterol that were used had manufac-
turer declared purities of only >70% and 65%, respectively, the con-
tent and composition of sterols in the synthetic steryl ferulates
were checked by GC after saponification followed by derivatization
to the trimethylsilyl ether, as described by Winkler-Moser et al.
(2013).
ferulate), the same procedure used for cholestanyl ferulate pro-
vided 4.8 g (89%) white solid. mp 161–162 °C; IR (cm ) 3436,
2
ꢁ
1
1
934, 2867, 2849, 1708, 1636, 1263, 1170, 1156, 1030; H NMR
(
CDCl ) d 0.70 (s, 3H), 0.88 (d, J = 2.2 Hz, 3H), 0.89 (d, J = 2.2 Hz,
3
3
2
H), 0.94 (d, J = 6.6 Hz, 3H), 0.96–1.05 (m, 3H), 1.07 (s, 3H), 1.08–
.08 (m, 23H), 2.42 (m, 2H), 3.93 (s, 3H), 4.76 (m, 1H), 5.41 (br
m, 1H), 5.93 (s, 1H), 6.29 (d, J = 15.9 Hz, 1H), 6.93 (d, J = 8.2 Hz,
1
H), 7.04 (d, J = 1.6 Hz 1H), 7.08 (m, 1H), 7.61 (d, J = 15.8 Hz, 1H);
C NMR (CDCl ) d 11.9, 18.7, 19.4, 21.1, 22.6, 22.8, 23.9, 24.3,
3
1
3
2
4
1
7.9, 28.0, 28.2, 31.9, 31.9, 35.8, 36.2, 36.6, 37.1, 38.3, 39.5, 39.8,
2.3, 50.0, 55.9, 56.2, 56.7, 73.9, 109.3, 114.7, 116.1, 122.7, 123.0,
27.1, 139.7, 144.5, 146.8, 147.9, 166.7. Analysis by TLC showed
2.2.4. Antioxidant activity assay
The antioxidant activity was determined using the 2,2 -
azinobis-(3-ethylbenzothiazoline-6-sulphonic
decolorization assay as described by Re et al. (1999). Trolox (6-
hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid) was used
0
+
ꢀ
one spot.
For synthesis of 3-O-(trans-4-feruloyl)-b-sitosterol (sitosteryl
ferulate), the same procedure used for cholestanyl ferulate pro-
vided 4.8 g (90%) white solid. H and C NMR data were consistent
with published literature values (Condo et al., 2001): mp 131–
acid)
(ABTS )
1
13
as the antioxidant standard and was tested in triplicate at 2
M, 10 M, and 15 M (final assay concentration). Steryl feru-
lates and ferulic acid were also tested in triplicate at 2 M, 5 M,
10 M, and 15 M. At these concentrations, the steryl ferulates
lM,
5
l
l
l
ꢁ
1
1
2
0
3
2
32 °C lit. 156–157 °C (Condo et al., 2001); IR (cm ) 3377, 2935,
l
l
867, 1701, 1633, 1593, 1267, 1157, 1030; 1H NMR (CDCl
.72 (s, 3H), 0.84 (d, 3H, J = 6.9 Hz), 0.85–0.89 (m, 6H), 0.95 (d,
H, J = 6.6 Hz), 0.96–1.05 (m, 3H), 1.06 (s, 3H), 1.07–2.08 (m,
4H), 2.42 (m, 2H), 3.93 (s, 3H), 4.76 (m, 1H), 5.41 (m, 1H), 5.92
) d
l
l
3
were soluble in ethanol. We compared results for several steryl fer-
ulates by first dissolving them in either dichloromethane or etha-
nol, but the results were the same using either solvent. In brief,
(
s, 1H), 6.29 (d, J = 15.9 Hz, 1H), 6.92 (d, J = 8.2 Hz, 1H), 7.04 (d,
J = 1.8 Hz, 1H), 7.06 (d, J = 8.0 Hz, 1H), 7.63 (d, J = 15.9 Hz, 1H);
10
ll of test compound or standard, dissolved in ethanol, was
pipetted into three cuvettes, while 10
l
l of ethanol was added to
1
3
ꢀ+
C NMR (CDCl
3
) d 11.9, 12.0, 18.8, 19.1, 19.4, 19.8, 21.1, 23.1,
another cuvette to serve as the control. One mL of the ABTS solu-
tion, prepared as described by Re et al. (1999), was added and the
cuvettes were placed into a Perkin Elmer Lambda 35 UV–vis spec-
trophotometer set to take automatic readings at 734 nm. Readings
were started exactly 30 s after the addition of the ABTSꢀ solution,
and were taken every 30 s for 5 min. The percentage inhibition of
absorption of each compound at 1 min and 5 min was calculated,
and the percentage inhibition at both time points was plotted as
a function of the concentration. The trolox equivalent antioxidant
capacity (TEAC) was calculated by dividing the slope of the least-
squares best-fitting line for the sample by the slope of the trolox
line.
2
3
1
4.3, 26.1, 27.9, 28.3, 29.2, 31.9, 31.9, 34.0, 36.2, 36.6, 37.1, 38.3,
9.8, 42.3, 45.9, 50.1, 55.9, 56.1, 56.7, 73.9, 109.3, 114.7, 116.1,
22.7, 123.0, 127.1, 139.7, 144.5, 146.8, 147.9, 166.7. Analysis by
+
TLC showed one spot.
For synthesis of 3-O-(trans-4-feruloyl)-8,24,(5-a)-cholestadien-
4
,4,14-a-trimethyl-3-b-ol (lanosteryl ferulate), the same procedure
used for cholestanyl ferulate provided 4.3 g (81%) white solid. mp
ꢁ1
1
1
3
1
1
67–171 °C; IR (cm ) 3350, 2948, 2872, 1676, 1635, 1592, 1254,
182, 1159, 1010; 1H NMR (CDCl
) d 0.72 (s, 3H), 0.86–0.89 (m,
3
H), 0.90 (s, 3H), 0.91–0.94 (m, 3H), 0.95 (s, 3H), 0.99 (s, 3H),
.05 (s, 3H), (m, 12H), 1.62 (s, 3H), 1.70 (s, 3H), 1.71–2.15 (m,
1H), 3.95 (s, 3H), 4.67 (m, 1H), 5.12 (m, 1H), 5.90 (s, 1H), 6.31
(
d, J = 15.9 Hz, 1H), 6.93 (d, J = 8.2 Hz, 1H), 7.05 (d, J = 1.7 Hz, 1H),
2.2.5. Oxidative stability index
7
1
2
3
1
1
.09 (m, 1H), 7.61 (d, J = 15.9 Hz, 1H); 13C NMR (CDCl
3
) d 15.8,
Soybean oils were prepared with steryl ferulate concentrations
6.7, 17.6, 18.2, 18.6, 18.7, 19.2, 21.0, 22.6, 22.8, 24.3, 24.3, 24.3,
4.9, 25.7, 26.4, 28.0, 28.0, 28.2, 28.2, 30.8, 31.0, 35.3, 36.3, 36.4,
6.5, 37.0, 38.1, 39.5, 44.5, 44.5, 49.8, 50.4, 50.6, 56.0, 80.8, 109.3,
14.7, 116.3, 123.0, 125.3, 127.1, 130.9, 134.3, 134.5, 144.3,
46.8, 147.9, 167.1. Analysis by TLC showed one spot.
ranging from 0.4413 to 8.85
using steryl ferulate stock solutions dissolved in chloroform. Oils
prepared with ferulic acid (0.441, 0.885, and 4.41 mol/g) were
lmol/g (ꢀ250
l
g/g to ꢀ5000
lg/g)
l
also prepared using stock solutions dissolved in ethanol. An appro-
priate volume of stock solution was added to amber vials and evap-
orated under a gentle stream of nitrogen. Oil was weighed into the
vials, and the vials were sonicated for 15 min, at room tempera-
ture, to dissolve the steryl ferulates, followed by vortexing for
10 s. The control soybean oil samples were sonicated and vortexed
in the same manner as the treatment oils. The OSI at 110 °C was
determined in duplicate following the AOCS Official Method Cd
12b-92 (1998). A Metrohm (Herisau, Switzerland) 743 Rancimat
with software control automatically controlled air flow and tem-
perature and calculated the OSI values based on induction time.
Steryl ferulates and ferulic acid were analysed at each concentra-
tion level together in duplicate, and a control soybean oil was
included with each set.
For synthesis of 3-O-(trans-4-feruloyl)-stigmastanol (stigmas-
tanyl ferulate), the same procedure used for cholestanyl ferulate
provided 4.6 g (86%) white solid. mp 156–157 °C; IR (cm ) 3350,
2
(
1
7
ꢁ1
1
931, 2866, 1701, 1676, 1633, 1592, 1257, 1158, 1010; H NMR
CDCl ) d 0.67 (s, 3H), 0.74–2.06 (m, 47H), 3.93 (s, 3H), 4.84 (m,
H), 5.91 (s, 1H), 6.28 (d, J = 15.9 Hz, 1H), 6.92 (d, J = 8.2 Hz, 1H),
3
1
3
.04 (d, J = 1.8 Hz, 1H), 7.07 (m, 1H), 7.60 (d, J = 15.9 Hz, 1H);
) d 12.0, 12.1, 12.3, 18.7, 19.1, 19.8, 21.2, 23.1, 24.2,
6.1, 27.6, 28.3, 28.7, 29.2, 32.0, 33.9, 34.2, 35.5, 36.2, 36.8, 40.0,
C
NMR (CDCl
3
2
4
1
2.6, 44.7, 45.9, 54.3, 55.9, 56.2, 56.4, 73.7, 109.3, 114.7, 116.2,
23.0, 127.1, 144.4, 146.8, 147.8, 166.8. Analysis by TLC showed
one spot.

9
6
J.K. Winkler-Moser et al. / Food Chemistry 169 (2015) 92–101
2.2.6. Frying studies
in intensity of the peaks are reported as percentage of the original
In previous frying and heating studies, steryl ferulates had
amount of the olefinic protons (ACH@CHA, multiplet, 5.30–
strong anti-polymerisation activity at a concentration of 0.5% (w/
w). Using the molecular weight of cholestanyl ferulate as the arbi-
5.44 ppm), bisallyllic protons (@CHACH
2.74–2.85 ppm) and allylic protons (@CHACH
2
ACH@, multiplet,
ACH A, multiplet,
2
2
trarily chosen standard, 0.5% is equal to 8.85
l
mol/g oil. The mass
1.98–2.13 ppm).
concentration of each steryl ferulate was adjusted so that each was
studied at the same molar concentration, thus the weight concen-
tration for the seven steryl ferulates ranged from 4.98 to 5.40 mg/
mL. Steryl ferulates were weighed into 10 mL volumetric flasks and
brought to volume with chloroform for a concentration of 20.0 mg/
mL. The appropriate volume of each test compound solution was
pipetted into 10 mL Pierce Reacti-therm glass vials and the solvent
was evaporated under a gentle stream of nitrogen. Soybean oil
Total polar compounds were determined according to a micro
method developed by Marquez-Ruiz, Jorge, Martin-Polvillo, and
Dobarganes (1996) and modified in our laboratory. This method
is a modification of the American Oil Chemists’ Society (AOCS)
official method Cd 20-91 (97) (1998). Micro methods utilising
SepPaks have been shown to give similar results to the standard
methods. Oils were separated into nonpolar and polar fractions
on a1g silica SepPak (Waters Corporation). The column was condi-
tioned with 10 mL of a 90:10 (v/v) petroleum ether to diethyl ether
(
6.5 g) was added. After filling the headspace with argon, the vials
were capped and stored at ꢁ20 °C for one to 5 days prior to the fry-
3
mixture, and then 1 mL of a 25 mg/mL oil solution in CHCl was
ing studies.
loaded onto the SepPak. The nonpolar fraction was eluted with
an additional 10 mL of the 90:10 (v/v) petroleum ether to diethyl
ether mixture, followed by 10 mL of a 50:50 (v/v) chloroform to
methanol mixture to elute the polar fraction. During method
development, separations were monitored by TLC to verify the sep-
aration of fractions. Fractions were dried under nitrogen and gentle
heat for several hours and weighed upon dryness. Experiments
were conducted in triplicate, and each sample was analysed in
duplicate.
The frying studies were conducted using a miniature frying pro-
tocol that was described in detail previously (Winkler-Moser et al.,
2
012). The frying apparatus limited the amount of treatments that
could be compared at one time, so a partially balanced, random-
ized incomplete block design was used where all seven treatments
were studied in triplicate frying studies completed over a series of
four frying experiments. After vials of oil were thawed, they were
placed in a Reacti-Vap aluminum block and heated to 180 °C using
a hot plate connected to a temperature controller (J-KEM Scientific,
St. Louis, MO). When the oil reached 70 °C, glass pipettes were used
to mix the oil to make sure that the steryl ferulates dissolved com-
pletely. Once the oil reached 180 °C, a zero-time sample was taken
Steryl ferulates in soybean oil at time 0 and during the frying
study were measured by HPLC as described by Winkler-Moser
et al. (2013).
(
500 mg), followed by the first fry cycle, which consisted of frying
2
.3. Statistical analysis
one 0.6 cm potato cube per vial for 2 min. Fry cycles were repeated
every hour for the next seven hours, and additional oil samples
(
ended after a total of 8 h at 180 °C. Oil samples were kept in 2 mL
capped vials, covered with argon, and frozen (ꢁ80 °C) until analy-
sis could be conducted.
For statistical analysis of the treatments on antioxidant activity,
500 mg) were taken after 2, 4, 6, and 8 h. The frying experiments
total polar compounds, and NMR analysis of double bonds, data
were imported to JMP (Statistical Analytical Systems, Cary, NC)
for analysis of variance (ANOVA) using a level of significance of
P < 0.05. Statistical differences between treatment means were
evaluated either by Fisher’s LSD or by the Tukey–Kramer HSD test
using a level of significance of P < 0.05. For analysis of statistical
significance of the formation of polymerised triacylglycerols
during frying, the mean percentage PTAGs were weighted by
2
.2.7. Analysis of frying oils
Polymerised triacylglycerols (PTAG) in the oil samples were
determined using the Association of Official Analytical Chemists
AOAC) Official Method 993.25 (2002) with the exceptions that
(
1
/variance, and the best fitting regression curves were determined
sample concentration was 5 mg/mL, methylene chloride was used
as solvent instead of tetrahydrofuran, and that an evaporative light
scattering detector (Shimadzu ELSD-LT II) was used rather than a
refractive index detector. Samples were injected into a Shimadzu
HPLC (model LC20AT) equipped with an autosampler, membrane
degasser, column heater (30 °C) and a size exclusion column (PLGel
using TableCurve 2D version 5.01 (Systat Software, San Jose, CA,
USA). Upper and lower 95% confidence intervals were determined
at each time point using the predicted PTAGs values for each treat-
ment. Treatment differences were considered significant (P < 0.05)
when there was no overlap of the confidence intervals. Microsoft
Excel 2007 (Redmond, WA, USA) was used to calculate means
and standard deviations and to construct best-fitting lines for anti-
oxidant assay data.
5
l
m, 100 Å pore size, 300 ꢂ 7.5 mm Agilent Technologies, Santa
Clara, CA). The ELSD was operated at a temperature of 40 °C with
the nebulizer gas (ultra-pure N ) pressure set to 3.5 Bar, and the
2
Gain set to 1. HPLC control, data collection and analysis were per-
formed by Shimadzu EZStart Chromatography Software Version
3. Results and discussion
7
.3. All samples were analysed in duplicate. Data were reported
as peak area %, which was the sum of dimer, trimer, and polymer
peaks divided by the total peak area including triacylglycerol
monomers.
Reactions at reactive sites of SBO triacylglycerols were moni-
tored during heating as previously described (Hwang, Winkler-
Moser, & Liu, 2012). In brief, 1H NMR spectra were acquired after
3.1. Synthesis of steryl ferulates
The previously reported method (Ebenezer, 1991) for the syn-
thesis of trans-4-O-acetylferulic acid requires solvent extraction
using a large amount of solvent (3 ꢂ 20 mL dichloromethane for
0.66 g product) because the product has low solubility in most
organic solvents. The solvent extraction process was also followed
by successive washing processes with 2 N HCl, water and brine,
which resulted in use of a large volume of the aqueous solution.
This procedure is cumbersome for a large-scale synthesis. A larger
scale synthesis (25 g) was reported (Condo et al., 2001), however, a
long reaction time (overnight) and four-times repeated addition of
solvent and evaporation (toluene, 4 ꢂ 50 mL) were involved. This
method does not have a purification step and, therefore, the purity
3
diluting 60 mg oil sample with CDCl as solvent using the NMR
equipment and software previously described. No measurable
hydrolysis affecting the peak intensity of the triacylglycerol back-
bone protons was observed up to 8 h of heating in a quantitative
experiment in which an exact amount (60 mg) of sample dissolved
in 0.5 mL solvent was used. Therefore, integration of peak area was
conducted with the triacylglycerol backbone protons as the stan-
dard peak (two double doublets, 4H, 4.06–4.36 ppm). The changes

J.K. Winkler-Moser et al. / Food Chemistry 169 (2015) 92–101
97
was not confirmed. Low purity of trans-4-O-acetylferulic acid may
result in a low yield for the next step in the synthesis. Therefore,
we developed a new synthetic procedure for trans-4-O-acetylferu-
lic acid that provides a high yield and high purity and that requires
a shorter time for better suitability for a large scale preparation. In
this new procedure, the reaction time is about 4 h, the excessive
acetic anhydride is hydrolysed with water, and pyridine is washed
with water after neutralization. For purification, impurities were
dissolved in methanol and removed by filtration.
The synthesis of trans-ferulolyl-b-sitostanol using the coupling
reaction of trans-4-O-acetylferulic acid and b-sitostanol catalyzed
by DCC was previously reported (Condo et al., 2001). This method
provided somewhat low yield (39–44%) for the coupling reaction
and starting materials, trans-4-O-acetylferulic acid and sterols
were not completely reacted. Furthermore, reaction time was
relatively long (18 h) and complete removal of the byproduct,
additional double bond at C24 in the side chain. Lanosteryl ferulate
had two sterol peaks; lanosterol made up 63.2% of the peak area,
while the other peak did not elute at the retention time of any
known standards, so the sample (as the trimethylsilylether) was
injected on a GC–MS for identification. The minor peak (36.8%) cor-
responded to 24-dihydrolanosterol, which does not have a double
bond at C24, but still has the main structural feature of interest, the
4,4 -dimethyl group. Sitostanyl ferulate was 97% sitostanol, cho-
lesteryl ferulate had 95% cholesterol, and cholestanyl ferulate had
98% cholestanol.
0
3.2. Antioxidant activity of steryl ferulates
The antioxidant activity of the synthesised steryl ferulates,
oryzanol, and ferulic acid were evaluated by the ABTSꢀ radical cat-
ion decolorization assay and in soybean oil by measuring the oxi-
+
ꢀ+
1
,3-dicyclohexylurea (DCU), took a long time (overnight) to be
dative stability index (OSI). In the ABTS
radical cation
precipitated out from THF. chemoenzymatic synthesis of
A
decolorization assay, ABTS is reacted with potassium persulfate
ꢀ+
phytosteryl ferulates was also recently reported by Tan and
Shahidi (2011) where vinyl ferulate was synthesised via a vinyl
interchange reaction of vinyl acetate with ferulic acid, followed
by lipase catalyzed synthesis of phytosteryl ferulates using a
to generate a stable ABTS radical which is bright green–blue in
colour and has an absorption maximum of 734 nm. The absorbance
ꢀ+
of the ABTS solution decreases when an electron is donated by an
antioxidant to quench the free radical; the % inhibition is propor-
tional to the concentration of the antioxidant. When the % inhibi-
tion is plotted as a function of the concentration and the slope of
the linear regression line is divided by the slope of the trolox linear
regression line, the trolox equivalent antioxidant capacity (TEAC),
1
:2 mol ratio of vinyl ferulate to phytosterols. Synthesis of the
vinyl ferulate intermediate took 12hrand resulted in a yield of
6%. Ten lipase sources were investigated and Candida rugosa
4
lipase was the only enzyme which catalyzed the esterification
of phytosteryl ferulates. A yield of 90% was reported, but the
reaction took 10 d. Thus, for this study a chemical synthesis route
was chosen in order to obtain gram scale quantities of pure phyt-
osteryl ferulates. In order to increase the yield and purity, some
of the existing procedures (Condo et al., 2001; Ebenezer, 1991)
were improved. In this new procedure, the reaction time was
reduced to 1.5 h, DCU was precipitated instantly by adding hex-
anes, column chromatography was used for purification instead
of the preparative liquid chromatography (PLC), and the final
yield was improved to 77–90%. The reported procedure was fol-
lowed for removal of the acetate protecting group (Condo et al.,
or the
lM concentration of sample that has the same activity as
1
lM trolox, can be calculated. As an example, a TEAC of 0.5 means
the compound has half of the radical scavenging capacity of trolox.
Table 1 shows the % inhibition after 5 min reaction for each com-
pound at each concentration, as well as the TEAC calculated at
5 min. Ferulic acid had the highest radical scavenging activity, with
a TEAC of 1.155. For ferulic acid and the steryl ferulates, the radical
scavenging activity increased between 1 and 5 min, indicating that
ꢀ
+
they react more slowly with the ABTS radical compared to trolox,
which reacts much more quickly and does not increase in antioxi-
dant activity much between 1 and 5 min. The % inhibition by feru-
2
001).
Since commercial phytosterol standards often contain impuri-
lic acid leveled off between 10 lM and 15 lM, so only the first
three concentrations were used to determine the slope and to cal-
culate the TEAC for ferulic acid. Others reported a slightly higher
TEAC for ferulic acid, from 1.69 to 1.90 (Re et al., 1999). The fact
that ferulic acid had much higher activity than the steryl ferulates
indicates that either solubility or steric hindrance prevented the
steryl ferulates from reacting as efficiently with the radical. This
assay has been used for both hydrophilic and lipophilic antioxi-
dants such as tocopherols and carotenoids (Re et al., 1999).
Furthermore, we tested oryzanol using both methanol and methy-
lene chloride as a solvent, and found that there was no difference
in activity. Also, since we found that activity of steryl ferulates
ties, and because the commercial sitosterol and lanosterol that
were used had manufacturer declared purities of >70% and 65%,
respectively, the content and composition of sterols in the syn-
thetic steryl ferulates were checked by GC. Sitosteryl ferulate was
composed of 80% (peak area percentage) sitosterol, while several
other sterol peaks were identified, including campesterol (8.7%),
campestanol (2.9%), stigmasterol (7.2%), and sitostanol (0.8%).
These contaminants were likely co-extracted from soybean oil
from which the sitosterol was extracted. Campesterol and stigmas-
terol also have a delta-5 double bond, but stigmasterol has an
Table 1
%
Inhibition of ABTSꢀ radical, and trolox equivalent antioxidant capacity (TEAC), for synthetic steryl ferulates and ferulic acid (n = 3).
+
%
2
Inhibition at 5 min (Ave. ± SD)
TEAC
lM
5
lM
10
l
M
15
l
M
Overall1
29.08b
(5 min)
Cholestanyl ferulate
Cholesteryl ferulate
Sitostanyl ferulate
Sitosteryl ferulate
Lanosteryl ferulate
Oryzanol
8.91 ± 0.81b
18.54 ± 1.74b
17.67 ± 0.35b
19.51 ± 1.86b
41.47 ± 0.98b
37.17 ± 1.55
35.01 ± 3.13d
35.15 ± 0.60
37.98 ± 2.11
31.67 ± 0.52
57.08 ± 0.65
48.49 ± 2.03c
45.41 ± 1.69d
43.70 ± 1.33d
43.13 ± 1.13
56.56 ± 1.75
0.887
0.829
0.762
0.795
1.094
0.867
1.155
bc
cd
b,c
7.75 ± 0.52
8.03 ± 1.97
26.75
26.33
bc
b,c
6.59 ± 0.41c
17.33 ± 2.23
14.71 ± 1.20
18.67 ± 0.65
43.79 ± 0.98
b
cd
d
25.31
28.70
c
bc
c
c
e
a
b
b
6.88 ± 2.04
7.77 ± 0.38
22.41 ± 1.37
bc
b
d
c
44.52 ± 0.43
61.02 ± 0.19
25.49
a
a
a
a
Ferulic acid
45.84
a,b,c,d,e
Mean values within the same column with different superscript letters are significantly different at P < 0.05 by Fisher’s LSD test.
See Section 2.2.4. to see how TEAC is calculated.
1
Standard error of the means was 1.05.

9
8
J.K. Winkler-Moser et al. / Food Chemistry 169 (2015) 92–101
Table 2
OSI (hr) at 110 °C, and OSI percentage of control soybean oil (in parentheses), for synthetic steryl ferulates, oryzanol, and ferulic acid added to soybean oil (n = 2).
OSI ± SD (% SBO)
Treatment
0.44
lmol/g
0.88
lmol/g
4.43
l
mol/g
8.85 lmol/g
Soybean oil (control)1
Cholestanyl ferulate
Cholesteryl ferulate
Sitostanyl ferulate
Sitosteryl ferulate
Lanosteryl ferulate
Oryzanol
7.89 ± 0.13
7.68 ± 0.13ab
8.61 ± 0.12a
8.60 ± 0.02a
a
cd
de
ab
e
bc
d
7.86 ± 0.01 (99.6)
7.75 ± 0.13 (98.2)
7.65 ± 0.01 (97.0)
7.83 ± 0.11 (99.2)
7.86 ± 0.01 (99.6)
7.92 ± 0.06 (100)
7.56 ± 0.01 (96.1)
7.95 ± 0.03 (104)
7.81 ± 0.11 (90.7)
7.44 ± 0.11 (86.5)
8.26 ± 0.01 (96.0)
7.38 ± 0.15 (85.7)
8.13 ± 0.04 (94.5)
8.39 ± 0.01 (97.5)
7.08 ± 2.06 (82.2)
7.35 ± 0.13 (85.5)
6.44 ± 0.05 (74.9)
7.32 ± 0.13 (85.2)
7.50 ± 0.06 (87.2)
5.622 (65.4)e
7.23 ± 0.03 (94.2)b
ab
bc
b
7.71 ± 0.23 (100)
7.27 ± 0.09 (94.7)
6.46 ± 0.17 (84.2)
7.61 ± 0.08 (99.2)
6.63 ± 0.08 (86.3)
b
c
bc
ab
bcde
ab
c
7.15 ± 0.01 (83.2)
c
3
Ferulic acid
ND
a,b,c,d,e
Mean values within the same column with different superscript letters are significantly different at P < 0.05 by Tukey’s HSD test. In columns with no superscripts,
ANOVA showed no significant effect of treatment on OSI.
1
Two different batches of soybean oil were used for the analysis at 0.44 and 0.88 lmol/g vs 4.43 and 8.85 lmol/g, explaining the difference in SBO OSI between the two
sets of data.
2
3
Due to machine error, only one sample was analysed.
Not determined, ferulic acid was not soluble at this concentration.
increased with concentration, which would not be expected if they
were insoluble at these concentrations, we concluded that the
lower activity is likely due to steric hindrance by the sterol moiety
interfering with the ability of the radical and the phenolic group to
react. However, as concentration increased, the % inhibition by ste-
ryl ferulates increased linearly and approached the activity of feru-
lic acid. Although at each concentration, there were some
significant differences in % inhibition between the steryl ferulates,
there was no clear trend for the radical scavenging activity based
on the sterol structural features. TEAC values indicated that lanos-
teryl ferulate had the highest antioxidant activity, but this value
appears to have been highly influenced by the % inhibition at
activity of water soluble phenolic compounds such as trolox and
ferulic acid giving them effectively lower antioxidant activity. This
makes it difficult to make direct comparisons between the lipo-
philic ORAC assay and other antioxidant assays using trolox as a
standard. Tan and Shahidi (2011) also demonstrated improved
antioxidant activity of the phytosteryl ferulates compared to feru-
lic acid in cooked ground pork stored at 4 °C, indicating that the
increased hydrophobicity of the phytosteryl ferulates increased
their activity in a high-fat food system.
Previous studies had shown antioxidant activity during frying
for steryl ferulates at an addition level of 5.0 mg/g in soybean oil
(Winkler-Moser et al., 2012), which is approximately 8.85 lmol/g
1
5
lM, whereas it had lower % inhibition at all of the other concen-
oil, so the antioxidant activity of phytosteryl ferulates was assessed
trations. Two-factor analysis of variance indicated that there was a
significant effect of sterol head group, but there was also a signif-
icant interaction between the concentration and the sterol head
group. Comparison of the mean % inhibition over all concentrations
indicated that cholestanyl ferulate had the best activity, but was
not significantly higher than lanosteryl ferulate, cholesteryl feru-
late or sitostanyl ferulate. However, cholestanyl ferulate was sig-
nificantly higher in overall antioxidant activity compared to both
oryzanol and sitosteryl ferulate. We have previously found (unpub-
lished results), that corn steryl ferulates and oryzanol had similar
antioxidant activities by the ABTS assay when evaluated at a con-
by the effect on the oxidative stability index of soybean oil at two
high level concentrations, 4.43 lmol/g and 8.85 lmol/g oil, which
for steryl ferulates corresponded to 2.5–2.7 mg/g and 4.9–5.3 mg/
g, as well as two lower concentrations levels, 0.44 and
0.89
to a typical concentration range for added antioxidants. Ferulic
acid was only assessed at 0.44 mol/g, 0.89 mol/g, and
4.3 mol/g (0.86 mg/g) because it was not soluble in the oil at
higher concentrations.
lmol/g (ꢀ0.25 and 0.5 mg/g, respectively), which corresponds
l
l
l
At the lowest concentration, ferulic acid and the steryl ferulates
had very little impact on the OSI of soybean oil (Table 2), and in
fact there was no significant treatment effect on the OSI. At
0.88 lmol/g, both ferulic acid and lanosteryl ferulate had a signif-
icant prooxidant effect on soybean oil OSI, while the other steryl
centration range from 10 to 50
was 0.600, which is lower than the current results because the
activity levels off above 15 M. Our results are somewhat in agree-
lM. The TEAC measured for both
l
ment with Nyström et al. (2005), who reported that ferulic acid
had significantly higher radical scavenging activity, using the diph-
enylpicrylhydrazyl (DPPH) radical, compared to sitosteryl ferulate,
cholesteryl ferulate, and steryl ferulate mixtures extracted from
wheat and rye. However, they also reported that these extracts,
which mostly contain sitostanyl- and campestanyl-ferulates, had
better radical scavenging activity compared to pure sitosteryl and
cholesteryl ferulate. In another study, Islam et al. (2009) reported
ferulate treatments had no significant effect on soybean oil OSI.
At a concentration of 4.43 lmol/g, it was difficult to completely
dissolve the ferulic acid, and the duplicate OSI values were 8.54 h
and 5.62 h, indicating an uneven distribution of the ferulic acid
in the oil. At this concentration, the steryl ferulates, except for sito-
stanyl ferulate and oryzanol, also had a significant prooxidant
effect on soybean oil, and at the highest concentration, all of the
steryl ferulates had a significant prooxidant effect. Lanosteryl feru-
late and cholesteryl ferulate had the greatest prooxidant effect at
similar DPPH scavenging activities for 1–60 lM cycloartenyl feru-
late, 24-methylenecycloartenyl ferulate, sitosteryl ferulate, and
ferulic acid. Tan and Shahidi (2011) reported improved antioxidant
activity for chemoenzymatically synthesised phytosteryl ferulates
8.85
l
mol/g, while cholesteryl and sitosteryl ferulate were the
mol/g. Since the steryl ferulates had very
most prooxidant at 4.4
l
little antioxidant effect in the concentration range that we tested,
no conclusions could be drawn about the effect of structure on
antioxidant activity in this system. The fact that steryl ferulates
with one or more sterol double bonds (oryzanol, cholesteryl-, sitos-
teryl-, and lanosteryl-ferulate) had the most prooxidant activity at
higher concentrations indicates that oxidation of the ring double
bond may accentuate the prooxidant activity due to oxidation of
the double bonds and interaction of resulting radicals with sur-
rounding triacylglycerols (Winkler & Warner, 2008). However,
compared to ferulic acid (ꢀ16
l
mol trolox equivalents/
lmol phyt-
osteryl ferulates vs ꢀ8.00
lmol troloxequivalents/ mol ferulic
l
acid) using a lipophilic oxygen radical absorbance capacity (ORAC)
assay. In this assay, lipophilic compounds are solubilised using ran-
domly methylated b-cyclodextrin (RMCD), which allows for mea-
surement of the antioxidant activity of lipophilic substances in
the otherwise aqueous environment. However, in our experience
(
unpublished data), the RMCD interferes with the antioxidant

J.K. Winkler-Moser et al. / Food Chemistry 169 (2015) 92–101
99
the prooxidant activity of ferulic acid itself indicates that the main
contributor is the ferulic acid moiety. When the concentration of
antioxidants that all have the same antioxidant function (i.e. chain
breaking phenolic antioxidants), are increased, the effect does not
always increase and in some cases, they can have a prooxidant
effect. This is because when concentrations exceed their optimal
levels, the phenolics and their corresponding phenolic radicals
can participate in side reactions which can actually promote oxida-
tion (Frankel, 2005, chap. 9; Yanishlieva, Kamal-Eldin, Marinova, &
Toneva, 2002).
Our results, which demonstrated little to no effect of steryl fer-
ulates on soybean oil OSI at lower concentrations, and loss of effi-
cacy/prooxidant activity at higher concentrations, are supported by
previous research (Wang, Hicks, & Moreau, 2002), where at
1 lmol/g oryzanol, ferulic acid, and sitostanyl-ferulate increased
.2
the oxidative stability of soybean oil at 90 °C by 7.5%, 10.8%, and
%, respectively, while at a concentration of 2.4 mol/g, the soy-
0
l
bean oil oxidative stability was increased by 1.7%, 4.8%, and 3.2%
for oryzanol, ferulic acid, and sitostanyl ferulate, respectively. In
another study, oryzanol, cycloartenyl ferulate, 24-methylenecyclo-
artanyl ferulate and ferulic acid (all at 0.4
OSI of 10% methyl linoleate in mineral oil 2.25–2.98-fold
Kikuzaki et al., 2002). The difference in results might be explained
lmol/g) increased the
(
by the highly oxidizable nature of the substrate, methyl linoleate,
compared to soybean oil, which has some protection from its
endogenous tocopherols. As an example, we previously demon-
strated that the addition of 0.5% and 1% of an antioxidant extract
from distillers dried grains containing steryl ferulates, tocopherols,
and tocotrienols had no effect on the OSI of soybean, sunflower,
and high-oleic sunflower oils, but increased the OSI of the same
oils in a dose dependent manner when the oils were stripped of
their native tocopherols (Winkler-Moser & Vaughn, 2009).
Fig. 1. Formation of polymerised triacylglycerols (PTAGs) during frying. See Table 3
for statistical analysis of PTAGs at each time point.
significantly after 6 and 8 h of frying, whereas oryzanol was only
significantly lower from the control at 8 h. The most remarkable
difference noted was between oryzanol and the synthetic steryl
ferulates. Lanosteryl ferulate prevented polymerisation better than
oryzanol, but it was significantly worse than the other steryl feru-
lates. Cholestanyl ferulate and sitostanyl ferulate prevented poly-
merisation in a similar fashion, and PTAGs were consistently
lower with these two treatments compared to their C5 unsaturated
counterparts, sitosteryl and cholesteryl ferulate. However, the dif-
ferences between the C5 saturated vs their unsaturated counter-
parts were not statistically significant except that at 8 h
sitosteryl ferulate was significantly higher in PTAGS compared to
sitostanyl- and cholestanyl-ferulate. These results indicate that
saturated phytosterols make slightly better antipolymerisation
agents compared to sterols with a double bond at C5. Lanosterol,
which has both the dimethyl group at C4, as well as double bonds
at C8 and C24, had activity in between the other sterols with zero
or one double bonds, and oryzanol. Since oryzanol is composed of a
mixture of sterols with a C5 double bond, sterols with a C4
dimethyl group, a C9,C19- cyclopropane group, and a C24 double
bond, these data indicate that the anti-polymerisation activity is
greatly impacted by the C9,C19 cyclopropane group, since this is
the major structural difference between oryzanol and lanosteryl
ferulate a the other synthetic steryl ferulates.
3.3. Effect of steryl ferulates in soybean oil during frying
One of the major products of oil oxidation and degradation dur-
ing frying is the formation of triacylglycerol polymers (PTAGs), as
well as a mixture of polar oxidised triacylglycerols, monoacylglyce-
rols and diacylglycerols, which are generally referred to as total
polar compounds (TPC). Measurement of these two products are
used worldwide as an indicator of oil quality during frying
(Gertz, 2000). Formation of PTAGs in the control soybean oil, as
well as in soybean oil with added steryl ferulates, are shown in
Fig. 1. Weighted regression equations (Table 3) were optimised
for each treatment and they all had high correlation coefficients
2
(
R P 0.922). For all treatments, PTAGs increased slowly between
0
h and 2 h, but after 4 h treatment differences in the rate of
increase in PTAGs were apparent. Treatments were compared by
non-overlap of 95% confidence intervals for the predicted % PTAGs
from the equations. The synthetic steryl ferulates lowered PTAGs
Table 3
2
Weighted regression equations for mean % PTAGs (Y-variable) as a function of time in hours (X-variable), adjusted R , treatment differences in predicted % PTAGs at each time
point, and change in % total polar compounds from 0 h to 8 h during frying (n = 3).
Treatment
Weighted regression equationsA
Adj. R2
Time
0
2
4
6
8
DTPC (%)
p
a
a
a
a
a
a
a
a
a
a
a
a
a
a
ab
ab
ab
ab
a
c
c
c
c
b
a
a
11.06 ± 2.66a
Control
ln(Y) = ꢁ4.867 + 2.309 *
ln(Y) = ꢁ6.156 + 2.127 *
ln(Y) = ꢁ8.356 + 3.112 *
ln(Y) = ꢁ6.326 + 2.174 *
ln(Y) = ꢁ7.392 + 2.850 *
ln(Y) = ꢁ7.343 + 2.982 *
ln(Y) = ꢁ4.335 + 2.042 *
X
X
X
X
X
X
X
0.922
0.929
0.964
0.962
0.992
0.987
0.972
0.0077
0.0021
0.0002
0.0018
0.0006
0.0006
0.2016
0.0430
0.0192
0.0387
0.0347
0.0439
0.2352
0.7798
0.1494
0.1187
0.1383
2.2020
0.3886
0.4807
0.3673
0.6630
0.9616
1.9480
5.2820
0.8702
1.5630
0.8371
1.9520
2.9770
p
p
p
p
p
p
e
cd
Cholestanyl
Cholesteryl
Sitostanyl
Sitosteryl
Lanosterol
Oryzanol
5.72 ± 0.55
7.31 ± 1.43
4.30 ± 1.31
5.77 ± 1.48
de
e
bc
d
0.1842b
d
c
cd
ab
ab
0.2517
0.7779
10.15 ± 0.89
a
4.2230b
10.48 ± 2.30a
0.0131
a,b,c,d,e
Values within a column followed by the same letter are not significantly different based on overlap of 95% confidence intervals from weighted regression equations for
predicted mean pTAGs, and based on Fisher’s LSD test for TPC.
Slope and intercept are significantly different from 0 at P 6 0.01 for all treatment equations.
D
A

1
00
J.K. Winkler-Moser et al. / Food Chemistry 169 (2015) 92–101
The antipolymerisation activity of the synthetic steryl ferulates
and 1.94% of the allylic double bond H protons. Oils with added
cholestanyl, cholesteryl, sitostanyl, and sitosteryl ferulate lost sig-
nificantly fewer allylic H protons, while only cholestanyl, choleste-
ryl, and sitosteryl lost significantly fewer olefinic protons, and only
cholestanyl lost significantly fewer bis-allylic protons. After 8 h
frying, neither lanosteryl ferulate nor oryzanol significantly pre-
vented the loss of olefinic, bis-allylic, or allylic double bonds. How-
ever, cholestanyl, cholesteryl, sitostanyl, and sitosteryl
significantly prevented the loss of olefinic and bis-allylic double
bond protons. For preventing loss of allylic double bond protons
after 8 h, the saturated steryl ferulates cholestanyl- and sitostanyl
ferulate were significantly better than their C5 unsaturated
counterparts.
Thus, three indicators for frying oil degradation all showed
superior antioxidant activity during frying for phytosteryl ferulates
composed of C4-desmethylsterols compared to C4-dimethylster-
ols. All three measurements indicated a trend showing that satu-
rated C4-desmethylsterols are slightly better antioxidants during
frying compared to those with a C5 double bond, however, the
differences were not great enough to be statistically significant.
During frying, lanosteryl ferulate was significantly less effective
than the C4-desmethylsteryl ferulates, indicating that either the
C4-dimethyl group, or the presence of two double bonds in the
structure of 65% of its components, affected its antioxidant activity.
Lanosteryl ferulate also showed a trend of better antioxidant activ-
ity compared to oryzanol, although this was only significant for
PTAG formation. Lanosteryl ferulate is similar in structure to the
major components in oryzanol, cycloartenyl ferulate and 24-meth-
ylene cycloartanyl ferulate, in that it has a 4,4-dimethyl group as
well as a double bond at C24. However, it also has an extra ring
double bond at C8. Since the minor components of oryzanol (36%
by weight) are made up of desmethylsterols similar in structure
to the sitosteryl and cholesteryl ferulate, it might be expected to
have antioxidant activity in between that of lanosteryl ferulate
and sitosteryl and cholesteryl ferulates, if the C9,C19 cyclopropane
group did not impact antioxidant activity. However, its consis-
tently lower antioxidant activity indicates that the C4-dimethyl
group and the C9,C19-cyclopropane group both have a significant
impact on its antioxidant activity during frying.
At this time, we cannot offer a complete explanation for why
structural differences in the sterol head group lend to varying anti-
oxidant activity during frying. The structural differences that we
examined in this study, including double bonds, dimethyl substitu-
tion, and the presence of a cyclopropane group, could affect the
stability of the phytosterol moiety to oxidation. In this study, the
average percentages of intact steryl ferulates remaining after 8 h
of frying were 74.6%, 78.5%, 78.7%, 88.0%, 88.2%, and 91.6% for oryz-
anol, lanosteryl-, sitosteryl-, cholesteryl-, sitostanyl-, and cholesta-
nyl ferulate, respectively. The percentages of steryl ferulates
remaining at 2, 4, 6, and 8 h were negatively correlated with PTAGs
(r = ꢁ.8159) and looking at 8 h alone, were negatively correlated
with PTAGs (r = ꢁ0.7805) and TPC (r = ꢁ0.6695). However, since
there was on average 75% of the oryzanol remaining and 78.7% of
sitosteryl ferulate remaining after 8 h, stability alone is unlikely
to explain the effectiveness of sitosteryl ferulate compared to oryz-
anol. A previous study showed that corn steryl ferulates were more
effective antioxidants compared to oryzanol and were also more
stable during frying in soybean oil (Winkler-Moser et al., 2012).
In a heating study at 180 °C, under conditions where soybean oil
oxidation progressed more rapidly, oryzanol was slightly more sta-
ble than corn steryl ferulates, but was still less effective of an anti-
oxidant (Winkler-Moser et al., 2013). In this study, the phytosterol
moieties were measured, and it was found that 50–75% of the phy-
tosterols remained intact, even though only 10–20% of the steryl
ferulates remained intact by the end of the study. The phytosterols
from oryzanol degraded to a greater extent than phytosterols from
compared to oryzanol in this study corresponded to previous
results showing higher antipolymerisation activity for corn steryl
ferulates, which are composed mainly of saturated sterols, com-
pared to oryzanol (Winkler-Moser et al., 2012; Winkler-Moser
et al., 2013), and also confirmed antipolymerisation activity of syn-
thetic sitostanyl ferulate in heated high-oleic sunflower oil
Nyström et al., 2007). We also demonstrated in a previous study
Winkler-Moser et al., 2012), that ferulic acid used at the same
molar concentration (8.85 lmol/g) had a prooxidant effect on soy-
(
(
bean oil polymerisation during frying, which is confirmed by the
OSI results (Table 2).
TPC were determined gravimetrically at the 0 h and 8 h time
points to compare to and verify the results from PTAG analysis. It
is important to note that for PTAG analysis, the area % reported is
not the same as weight percentage that is reported for TPC, due
to differences in the sensitivity of the evaporative light scattering
detector for monomeric vs polymeric triacylglycerol peaks. The
overall change in TPC from 0 to 8 h corresponded well with the
PTAG analysis (Table 3). Cholestanyl-, sitostanyl-, cholesteryl-,
and sitosteryl ferulate all prevented the formation of total polar
compounds after 8 h compared to the control SBO. However, oryz-
anol and lanosteryl ferulate did not significantly prevent formation
of total polar compounds, although lanosteryl ferulate showed a
similar trend to the PTAG formation in that it prevented TPC for-
mation slightly better than oryzanol.
During frying, the fatty acids in oils, especially polyunsaturated
fatty acids, undergo both oxidative and thermal reactions which
produce numerous compounds besides the polymerised triacylgly-
cerols measured previously, including aldehydes, ketones, and
short chain volatile compounds. These reactions result in a loss
of double bonds, which can be monitored by 1H NMR measure-
ment of the peak intensities of olefinic protons (ACH@CHA, multi-
plet, 5.39–5.45 ppm), bisallylic protons (@CHACH
2
ACH@,
A,
multiplet, 2.71–2.85 ppm), and allylic protons (@CHACH ACH
2
2
multiplet, 1.97–2.13 ppm) (Hwang et al., 2012; Wanasundara,
Shahidi, & Jablonski, 1995). Compared to analysis of fatty acid com-
position, this method is a more sensitive measure of double bond
loss, but it is less specific in that it does not specify which fatty
acids are affected by the loss of double bonds. As would be
expected, the bis-allylic protons were lost the fastest, followed
by the olefinic and allylic protons (Table 4). After 4 h, the control
SBO lost an average of 2.57% of the olefinic, 3.93% of bis-allylic,
Table 4
NMR analysis showing the mean percentage, compared to the 0 h time point in each
frying study, of olefinic, bis-allylic, and allylic double bond H protons remaining in
SBO during frying (n = 3).
Treatment
% Olefinic
% Bis-allylic
% Allylic
4
h
bc
a
98.06cd
Control-SBO
97.43cd
96.07
98.55
98.29
98.15
98.30
97.84
96.32
ab
ab
Cholestanyl ferulate
Cholesteryl ferulate
Sitostanyl ferulate
Sitosteryl ferulate
Lanosteryl ferulate
Oryzanol
99.08
99.49
99.04
99.27
99.37
99.47
99.32
a
ab
ab
ab
ab
abc
a
abc
ab
ab
ab
99.37
99.07
abc
abc
98.80
bcd
bcd
97.66
98.24
8
h
cd
cd
a
cd
Control-SBO
93.52
90.33
95.89
94.56
96.01
94.46
92.76
91.45
95.11
98.70
97.44
98.11
97.13
96.45
95.75
a
a
Cholestanyl ferulate
Cholesteryl ferulate
Sitostanyl ferulate
Sitosteryl ferulate
Lanosteryl ferulate
Oryzanol
97.03
96.39
97.37
95.97
95.18
94.26
ab
ab
a
abc
a
ab
ab
ab
abc
bcd
abc
abc
abcd
bcd
bcd
a,b,c,dMean values within the same column with different superscript letters are
significantly different at P < 0.05 by Tukey’s HSD test.

J.K. Winkler-Moser et al. / Food Chemistry 169 (2015) 92–101
101
corn steryl ferulates, so it is possible that oxidation products from
the oxidation of oryzanol and lanosteryl phytosterols interacted
with soybean triacylglycerols to have a prooxidant effect that
counteracted, or interfered, with their antioxidant effect.
Hwang, H.-S., Winkler-Moser, J. K., & Liu, S. X. (2012). Structural effects of lignans
and sesamol on polymerization of soybean oil at frying temperature. Journal of
the American Oil Chemists’ Society, 89, 1067–1076.
Islam, M. S., Yoshida, H., Matsuki, N., Ono, K., Nagasaka, R., Ushio, H., et al. (2009).
Antioxidant, free radical-scavenging, and NF-jB-inhibitory activities of
phytosteryl ferulates: Structure–activity studies. Journal of Pharmacological
Sciences, 111, 328–337.
Another possibility is that ferulic acid or oxidised ferulic acid
released from steryl ferulates by hydrolysis may be acting as pro-
oxidants as well. Ferulic acid acted as a prooxidant in the present
study by OSI analysis, and in a previous frying study (Winkler-
Moser et al., 2012), showed prooxidant activity and accelerated
the degradation of tocopherols. Unfortunately, our methods for
steryl ferulate analysis only indicate the levels of intact steryl fer-
ulates remaining, but do not explain if the loss is due to hydrolysis
or to oxidation or a combination of both. Thus, we need to explore
these and other possible explanations in order to understand the
differences observed in antioxidant activity during frying.
Overall, the results from this study indicate that despite having
similar antioxidant activities when compared using an in vitro anti-
oxidant activity assay, as well as by analysis of the effects on OSI in
soybean oil at 110 °C, steryl ferulates are dramatically affected by
the structure of the sterol group when comparing their antioxidant
activity in frying oils. These results indicate that saturated sterols
are slightly better, within the confines of our model frying experi-
ment, compared to more typical sterols which have a double bond
at C5. In addition, both types (saturated and C5-unsaturated ster-
ols) are superior to sterols such as those in oryzanol that have a
C4 dimethyl group, and a C9,C19 cyclopropane group. At least
within the parameters of this study, the 9,19 cyclopropane group
appeared to have a major impact on steryl ferulate antioxidant
activity during frying. Thermal and oxidative stability of the phy-
tosterol group may play a role on the antioxidant activity of the
steryl ferulates, but it may not have been the major factor involved
Juliano, C., Cossu, M., Alamanni, M. C., & Piu, L. (2005). Antioxidant activity of
gamma-oryzanol: Mechanism of action and its effect on oxidative stability of
pharmaceutical oils. International Journal of Pharmaceuticals, 299, 146–154.
Kikuzaki, H., Hisamoto, M., Hirose, K., Akiyama, K.,
& Taniguchi, H. (2002).
Antioxidant properties of ferulic acid and its related compounds. Journal of
Agricultural and Food Chemistry, 50, 2161–2168.
Marquez-Ruiz, G., Jorge, N., Martin-Polvillo, M., & Dobarganes, M. C. (1996). Rapid
quantitative determination of polar compounds in fats and oils by solid-phase
extraction and size-exclusion chromatography using monostearin as internal
standard. Journal of Chromatography A, 749, 55–60.
Marat, K. (2011). SpinWorks, version 3.1.8 beta 4. Manitoba, Canada: University of
Manitoba.
Moreau, R. A., Singh, V., Nuñez, A., & Hicks, K. B. (2000). Phytosterols in the aleurone
layer of corn kernels. Biochemistry Society Transactions, 28, 803–806.
Norton, R. A. (1995). Quantitation of steryl ferulate and p-coumarate esters from
corn and rice. Lipids, 30, 269–274.
Nyström, L., Mäkinen, M., Lampi, A.-M., & Piironen, V. (2005). Antioxidant activity of
steryl ferulate extracts from rye and wheat bran. Journal of Agricultural and Food
Chemistry, 53, 2503–2510.
Nyström, L., Achrenius, T., Lampi, A.-M., Moreau, R. A., & Piironen, V. (2007). A
comparison of the antioxidant properties of steryl ferulates with tocopherol at
high temperatures. Food Chemistry, 101, 947–954.
Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M., & Rice-Evans, C. (1999).
Antioxidant activity applying an improved ABTS radical cation decolorization
assay. Free Radical Biology Medicine, 26, 1231–1237.
Seitz, L. M. (1989). Stanol and sterol esters of ferulic and p-coumaric acids in wheat,
corn, rye, and triticale. Journal of Agricultural and Food Chemistry, 37, 662–667.
Tan, Z., & Shahidi, F. (2011). Chemoenzymatic synthesis of phytosteryl ferulates and
evaluation of their antioxidant activity. Journal of Agricultural and Food
Chemistry, 59, 12375–12383.
Shopova, N., & Milkova, T. (2000). Thermal decomposition of cholesteryl esters of
cynnamic acid derivatives and their effect on the a-tetralylhydroperoxyde free-
radical-induced decomposition. Thermochimica Acta, 356, 101–107.
Suh, M.-H., Yoo, S.-H., & Lee, H. G. (2007). Antioxidative activity and structural
(Winkler-Moser et al., 2012; Winkler-Moser et al., 2013). Future
stability of microencapsulated c-oryzanol in heat-treated lards. Food Chemistry,
100, 1065–1070.
studies will focus on determining the reason for differences in anti-
oxidant activity during frying, as well as to understand why steryl
ferulates are not especially strong antioxidants in vitro, nor as
assessed by OSI at 110 °C, but are very effective during frying.
Wanasundara, U. N., Shahidi, F., & Jablonski, C. R. (1995). Comparison of standard
and NMR methodologies for assessment of oxidative stability of canola and
soybean oils. Food Chemistry, 52, 249–253.
Wang, T., Hicks, K. B., & Moreau, R. (2002). Antioxidant activity of phytosterols,
oryzanol, and other phytosterol conjugates. Journal of the American Oil Chemists’
Society, 79, 1201–1206.
Winkler, J. K., & Warner, K. (2008). The effect of phytosterol concentration on
oxidative stability and thermal polymerization of heated oils. European Journal
of Lipid Science & Technology, 110, 455–464.
Winkler-Moser, J. K., & Vaughn, S. F. (2009). Antioxidant activity of phytochemicals
from distillers dried grain oil. Journal of the American Oil Chemists’ Society, 86,
1073–1082.
Acknowledgements
We would like to acknowledge Julie Anderson, Lynne Copes,
Kathy Rennick, Patrick Schneider, and Karl Vermillion for their
excellent technical expertise. We are also grateful to Bryan Moser
for generation of structures for Scheme 1.
Winkler-Moser, J. K., Rennick, K. A., Palmquist, D. A., Berhow, M. A., & Vaughn, S. F.
(
2012). Comparison of the impact of c-oryzanol and corn steryl ferulates on the
polymerization of soybean oil during frying. Journal of the American Oil Chemists’
Society, 89, 243–252.
References
Winkler-Moser, J. K., Rennick, K. A., Hwang, H.-S., Berhow, M. A., & Vaughn, S. F.
(2013). Effect of tocopherols on the anti-polymerization activity of oryzanol and
corn steryl ferulates in soybean oil. Journal of the American Oil Chemists’ Society,
90, 1351–1358.
American Oil Chemists’ Society (1998). Official methods and recommended practices
of the American Oil Chemists’ Society (5th ed.). Urbana, IL: AOCS Press.
Association of Official Analytical Chemists (2002). Official Methods of Analysis of
AOAC International (17th ed.). AOAC International: Gaithersburg, MD.
Condo, A. M., Jr., Baker, D. C., Moreau, R. A., & Hicks, K. B. (2001). Improved method
for the synthesis of trans-feruloyl-b-sitostanol. Journal of Agricultural and Food
Chemistry, 49, 4961–4964.
Xu, Z., & Godber, J. S. (1999). Purification and identification of components of c-
oryzanol in rice bran oil. Journal of Agricultural and Food Chemistry, 47,
2724–2728.
Xu, Z., & Godber, J. S. (2001). Antioxidant activities of major components of
c-
Ebenezer, W. J. (1991). Colabomycin co-metabolites. Synthesis of 2880-II,
a
oryzanol from rice bran using a linoleic acid model. Journal of the American Oil
Chemists’ Society, 78, 645–649.
Yanishlieva, N. V., Kamal-Eldin, A., Marinova, E. M., & Toneva, A. G. (2002). Kinetic
metabolite related to ferulic acid. Synthetic Communications, 21, 351–358.
Frankel, E. N. (2005). Lipid oxidation (2nd ed.). Bridgwater, England: The Oil Press P.J.
Barnes & Associates.
Gertz, C. (2000). Chemical and physical parameters as quality indicators of used
frying fats. European Journal of Lipid Science & Technology, 102, 566–572.
study of the antioxidant activity of
a- and c-TOHs in sunflower and soybean
oil triacylglycerols. European Journal of Lipid Science and Technology, 104,
262–270.