Synthesis and search for 3β,3′β-disteryl ethers after high-temperature treatment of sterol-rich samples

Article abstract of DOI:10.1016/j.foodchem.2020.127132

It has been proven that at increased temperature, sterols can undergo various chemical reactions e.g., oxidation, dehydrogenation, dehydration and polymerisation. The objectives of this study are to prove the existence of dimers and to quantitatively analyse the dimers (3β,3′β-disteryl ethers). Sterol-rich samples were heated at 180 °C, 200 °C and 220 °C for 1 to 5 h. Quantitative analyses of the 3β,3′β-disteryl ethers were conducted using liquid extraction, solid-phase extraction and gas chromatography coupled with mass spectrometry. Additionally, for the analyses, suitable standards were synthetized from native sterols. To identify the mechanism of 3β,3′β-disteryl ether formation at high temperatures, an attempt was made to use the proposed synthesis method. Additionally, due to the association of sterols and sterol derivatives with atherosclerosis, preliminary studies with synthetized 3β,3′β-disteryl ethers on endothelial cells were conducted.

Full text of DOI:10.1016/j.foodchem.2020.127132

FoodChemistry329(2020)127132
Contents lists available at ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Synthesis and search for 3β,3′β-disteryl ethers after high-temperature
treatment of sterol-rich samples
Adam Zmysłowski⁎, Jerzy Sitkowski, Katarzyna Bus, Karol Ofiara, Arkadiusz Szterk
National Medicines Institute, 30/34 Chełmska, 00-725 Warsaw, Poland
A R T I C L E I N F O
A B S T R A C T
Keywords:
It has been proven that at increased temperature, sterols can undergo various chemical reactions e.g., oxidation,
dehydrogenation, dehydration and polymerisation. The objectives of this study are to prove the existence of
dimers and to quantitatively analyse the dimers (3β,3′β-disteryl ethers). Sterol-rich samples were heated at
180 °C, 200 °C and 220 °C for 1 to 5 h. Quantitative analyses of the 3β,3′β-disteryl ethers were conducted using
liquid extraction, solid-phase extraction and gas chromatography coupled with mass spectrometry. Additionally,
for the analyses, suitable standards were synthetized from native sterols. To identify the mechanism of 3β,3′β-
disteryl ether formation at high temperatures, an attempt was made to use the proposed synthesis method.
Additionally, due to the association of sterols and sterol derivatives with atherosclerosis, preliminary studies
with synthetized 3β,3′β-disteryl ethers on endothelial cells were conducted.
Cholesterol
Sitosterol
Stigmasterol
3β,3′β-disteryl ether
Gas chromatography
Mass spectrometry
1. Introduction
Aladedunye, & Hazendonk, 2014; Struijs, Lampi, Ollilainen, & Piironen,
2010). After incubation for 3 h at 180 °C, 20% of the initially present
Sterols of both vegetable and animal origin are compounds that are
relatively easily oxidized in autoxidation reactions. Radical or reactive
forms of oxygen play a major role in reactions of this type. Sterol oxi-
dation products are various chemical compounds that contain addi-
tional oxygen groups such as alcohol, ketone or ether. The formation of
oxysterols/oxyphytosterols has been studied extensively (Derewiaka &
Obiedziński, 2010; Schött & Lütjohann, 2015; Szterk & Pakuła, 2016).
Various cholesterol oxidation products: 7-ketocholesterol (7-kCh), 7-
hydroxycholesterol (α and β isomer), 5,6-epoxycholesterol (α and β
isomer) and cholestane-3β,5,6β-triol have been proven to exhibit cy-
totoxicity as well as apoptotic and pro-inflammatory effects on human
cells (Kulig, Cwiklik, Jurkiewicz, Rog, & Vattulainen, 2016). Ad-
ditionally, after the discovery of the presence of oxysterol in athero-
sclerotic plaques in humans, their pro-atherosclerotic properties were
postulated (Zmysłowski & Szterk, 2017, 2019). However, the formation
of sterol oxidation products via oxidation is only one type of reaction
that sterols can undergo at high temperatures. The reactions that occur
above 180 °C are more diverse. In addition to oxidation reactions, de-
hydrogenation and dehydration reactions can occur (Chien, Wang, &
Chen, 1998). It has also been proven that various sterols can be con-
catenated if subjected to high-temperature treatment, which results in
polymeric structures (Derewiaka & Molińska (née Sosińska), 2015;
sterols are converted into various dimers (Lampi, Kemmo, Mäkelä,
Heikkinen, & Piironen, 2009; Menéndez-Carreño, Ansorena, Astiasarán,
Piironen, & Lampi, 2010). Cholesterol transformations during heat
treatment were studied by Derewiaka et al. (Derewiaka & Molińska (née
Sosińska), 2015). The use of higher temperatures, namely, 180 °C and
220 °C, led to the formation of polymers and other products, such as
cholestadienes and fragmented cholesterol molecules. Unfortunately,
the structures of the analysed cholesterol polymers were not confirmed.
Struijs et al. demonstrated the formation of stigmasterol polymers after
heating at 180 °C (Struijs et al., 2010). Based on molecular mass ob-
tained by the high-resolution mass spectrometry it was suggested that
one of the found dimers consisted of two stigmasterol moieties linked
with ether bond, which could correspond to the structure of 3β,3′β-
distigmasteryl ether. A similar investigation was published in the same
year by Rudzińska et al., which described the thermal oxidation of β-
sitosterol and the production of dimers, trimers, and tetramers, but no
results were presented on structural confirmation of the found poly-
mers. (Rudzińska et al., 2009). Such research was done by Sosinska
et al. (2013), which confirmed using combination of NMR, APCI/MS, IR
and Raman spectroscopy, that the most abundant dimeric product of
thermo-oxidation of sitosterol was 3β,3′β-disitosteryl ether (Sosińska,
Przybylski, Hazendonk, Zhao, & Curtis, 2013). However, those reac-
tions were studied only for heated sterols standards. Thus, a study
Rudzińska, Przybylski,
& Wąsowicz, 2009; Sosińska, Przybylski,
⁎ Corresponding author.
E-mail address: a.zmyslowski@nil.gov.pl (A. Zmysłowski).
https://doi.org/10.1016/j.foodchem.2020.127132
Received 21 March 2020; Received in revised form 20 May 2020; Accepted 22 May 2020
Availableonline23May2020
0308-8146/©2020ElsevierLtd.Allrightsreserved.

A. Zmysłowski, et al.
FoodChemistry329(2020)127132
should be conducted to determine whether such compounds can be
formed in food or biological samples. Very little information is available
on the properties of the dimers, such as their bioavailability and bio-
logical activity. Thus, it is desirable to synthesize such compounds to
the search for these compounds in biological matrices using spectro-
metric methods and to evaluate their biological properties.
2.3. Standard and sample preparation
2.3.1. Working solutions and calibration curve for the GC–MS method
The concentrations of 3β,3′β-disteryl ether were determined using a
standard calibration curve with synthesized isotope-labelled standards
(13, 14, 15). Isotope-labelled standards were dissolved in n-Hep to a
stock concentration of 1.0 mg/mL. A working solution with a con-
centration of 1.0 µg/mL for internal standards (ISs) was prepared from
the stock solution. Non-isotope-labelled standards were dissolved in n-
Hep to a stock concentration of 1.0 mg/mL. A working solution was
prepared from the 3β,3′β-disteryl ether stock solutions, which con-
tained a mixture of ethers (except for the isotope-labelled standards) at
a concentration of 1.2 μg/mL. Subsequent working solutions were
prepared via suitable dilutions of the stock solution (1.2 μg/mL) to the
desired concentrations. For calibration solutions, the following 3β,3′β-
disteryl ether concentrations were used: 500, 250, 100, 75, 50, 25 and
10 ng/mL. To these solutions, isotope-labelled standards (100 ng/mL)
were always added.
The objective of this study is to evaluate the possibility of formation
of sterols dimers – 3β,3′β-disteryl ethers during thermal treatment at
temperatures that are typical for food processing in a food matrix.
Hence, an attempt was made to prepare 3β,3′β-disteryl ethers, which
could be used as standards for an analytical method and as substances
for biological studies. In addition, a probable formation mechanism was
proposed based on the formation of various dimeric products.
Additionally, due to the association of sterols and sterol derivatives
with atherosclerosis, preliminary studies with synthetized 3β,3′β-dis-
teryl ethers on endothelial cells were conducted.
2. Materials and methods
2.1. Materials
2.3.2. Sample preparation of biological origin samples
The samples were chosen based on their high native sterol con-
centrations (Derina et al., 2018; Verleyen et al., 2002). Butter and cod-
liver oil due to high content of cholesterol were selected for analysing
the 3β,3′β-dicholesteryl ether, crude rapeseed oil due to high content of
β-sitosterol was selected for analysing the 3β,3′β-disitosteryl ether and
crude corn oil due to high content of β-sitosterol and stigmasterol was
selected for analysing the 3β,3′β-disitosteryl ether and the 3β,3′β-dis-
tigmasteryl ether. Prior to heating, the butter was clarified by melting it
at low temperature (60 °C) and removing the undissolved sediment.
The prepared samples were heated in open glass vials for the
durations and at the temperatures that are specified in Table 1 in an
oven. The samples were taken out from the oven and were allowed to
cool in room temperature, after which they were subjected to saponi-
fication. A portion of each sample (1.0 g in the cases of rapeseed oil,
corn oil and cod-liver oil and 0.25 g in case of butter) after the heat
treatment was mixed with 15 mL 1 M KOH in MeOH and left for 24 h in
the dark. Next, all saponificated samples were extracted at least twice
with 20 mL of n-Hex each, and the solvent was evaporated using a
rotavapor. The samples were redissolved in 5 mL of n-Hex and extracted
at least 3 times with 2 mL of MeOH. After the extraction, the solvent
was evaporated in a stream of nitrogen and redissolved in 1 mL of n-
Hex, and the samples were subjected to purification via solid-phase
extraction.
2.1.1. Chemical materials
All solvents for synthesis were of HPLC-grade quality. Isopropanol
(IPA), tetrahydrofuran (THF), dichloromethane (CH2Cl2), chloroform
(CHCl3), n-hexane (n-Hex), n-heptane (n-Hep), methanol (MeOH), die-
thyl ether (Et2O), ethyl acetate (EtOAc), acetic anhydride (Ac2O), me-
thyl tert-butyl ether (MTBE), dimethylformamide (DMF), dimethyl
sulfoxide (DMSO), pyridine and acetone, along with cyclohexane (c-
Hex) for gas chromatography MS SupraSolv®, were purchased from
Merck Millipore. All reagents were purchased from Aldrich, except
stigmasterol (Cayman Chemicals). The deuterated chloroform‑d1 (99.95
atom % D) was obtained from Dr Glaser at AG Basel. Rapeseed oil
(crude, cold-pressed), corn oil (crude, cold-pressed) and butter (based
on the label: 82% fat content) were purchased from a local market. Cod-
liver oil was purchased from a local pharmacy. Thin-layer chromato-
graphy (TLC) was conducted on precoated silica gel plates (Merck 60
PF254). The solid phase extraction (SPE) was done using Stata silica SI-
1 cartridges (500 mg, 3 mL, 55 µm particle size, 70 Å pore size) from
Phenomenex (Torrance, CA, USA). Visualization of compounds on TLC
plates was realized via ultraviolet (UV) (254 nm) light detection and/or
iodine staining. The melting points (mp.) are uncorrected and were
recorded on a capillary melting point apparatus. Dry column vacuum
chromatography (DCVC) was performed using Merck Silica gel 60
(Pedersen & Rosenbohm, 2001). Montmorillonite K 10 (MK10) was
dried for 2 h at 125˚C before the reaction. Each compound was obtained
in pure crystal form and was stored in 4 °C.
2.3.3. Solid phase extraction (SPE) for the determination of 3β,3′β-disteryl
ethers
SPE cartridges were conditioned with 5 mL of n-Hex. On SPE, 1 mL
of the n-Hex phase was applied. The cartridge was rinsed with 5 mL of
n-Hex. 3β,3′β-Disteryl ether was eluted with 10 mL of n-Hex:CH2Cl2
10:1 (v/v) solution. The solvent was evaporated under a stream of ni-
trogen at 30 °C. The residue was dissolved in 1 mL n-Hep.
2.1.2. Biological materials
Endothelial cells TeloHAEC were purchased from ATCC. The en-
dothelial cell growth medium for the in vitro culture was purchased
from PAN-Biotech GmbH. The medium contains epidermal growth
factor (EGF), fibroblast growth factor 2 (FGF-2), vascular endothelial
growth factor (VEGF), vitamin C, insulin-like growth factor-1 (R3-IGF-
1), 3% fetal bovine serum (FBS), gentamycin/amphotericin, hydro-
cortisone, and heparin. Trypsin TrypLe Express was purchased from
Thermo Scientific. The MTT test was purchased from Sigma-Aldrich.
1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) and 1,2-Distearoyl-
rac-glycero-3-methylpolyoxyethylene (DSG-PEG 2000) were acquired
from Avanti Polar.
2.4. Chromatographic methods
2.4.1. Preparation of the silver impregnated silica column
The Zorbax Si column of dimensions 250 × 4.6 mm 5 µm was used
for impregnation of silver ions. The column was flushed with 10 vo-
lumes of IPA and, subsequently, with 10 volumes of water. After
column equilibration with water, it was subjected to a 200 mL solution
of 5% AgNO3 with a 0.5 mL/min flow rate. After the solution was
consumed, the column was flushed with 2 volumes of water and, sub-
sequently, with IPA. The lines of the HPLC solvent carrier were flushed
with IPA to remove all water, and the column was flushed with n-Hep
and stored in the same solvent prior to use.
2.2. Chemical synthesis
The synthesis methods of all used compounds (1–33) along with the
1H and 13C NMR data can be found in the Supplementary Data. The
synthesis route is presented in Fig. 1.
2

A. Zmysłowski, et al.
FoodChemistry329(2020)127132
Fig. 1. Synthesis of sterol derivatives.
Table 1
Measurements of the 3β,3′β-disteryl ethers [µg/g] that are formed during thermal processing of sterol in various matrices at various temperatures. a,b,c)Values within
the same temperature of processing with different letters are significantly different (p < 0.05).
Food matrix
Contents of 3β,3′β-disteryl ethers [µg/g]
(n = 3)
Time [h]
Temperature
180 °C
200 °C
< LOD
220 °C
0.041
dCh
Cod-liver oil
Butter
1
2
3
4
5
1
2
3
4
5
< LOD
< LOD
0.023a
0.008b
0.004b
0.003b
0.007b
0.060a
0.094b
0.064b
0.053b
0.079a
0.006
0.020
0.030
0.023
0.199
0.240
0.254
0.331
0.325
0.009a
0.006b
0.013b
0.007b
0.029a
0.051a
0.029b
0.056c
0.052c
0.020
0.016
0.015
0.016
0.374
0.231
0.234
0.257
0.165
0.019
0.018
0.021
0.116
0.134
0.150
0.210
0.200
0.005a
0.010a
0.005a
0.015a
0.018a
0.026a
0.029b
0.025b
dStig
dSito
dStig
dSito
dStig
dSito
Rapeseed oil
Corn oil
1
2
3
4
5
1
2
3
4
5
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
0.043
0.022
0.010a
0.008b
0.040
0.005a
0.015a
0.005a
< LOD
0.049
0.044
< LOD
< LOD
0.040
0.004
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
0.030
0.010
0.012a
0.034
0.017
0.027a
0.014b
0.026
0.028
0.014a
0.007b
0.029a
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
< LOD
3

A. Zmysłowski, et al.
FoodChemistry329(2020)127132
2.4.2. HPLC method for 3β,3′β-disteryl ether synthesis optimization
The synthesis optimization was conducted using a Thermo Scientific
Accela Model 430 bar liquid chromatograph, an autosampler (10 °C), a
thermostat for the column (25 °C), and a diode array detector (DAD,
210 nm). 3β,3′β-Dicholesteryl ether was separated on Zorbax Si
250x4.6 mm 5 µm using n-Hep as the mobile phase. The sample was
dissolved in 1 mL of n-Hep after the evaporation of the reaction solvent.
The injection volume was 5 µL.
system was applied: 0–2 min 0% B, 2–10 min 0–60% B, 10–12 min 60%
B, and 12–15 min 0% B (balancing the column until the initial condi-
tions are restored). Chromatographic separation was conducted with a
constant flow of the mobile phase (400 µL/min) at a temperature of
35 °C.
2.5. NMR spectroscopy
The NMR spectra were recorded at 298 K by either a Varian
VNMRS-500 or Varian VNMRS-600 spectrometer that was equipped
with a 5-mm Z-SPEC Nalorac IDG 500-5HT gradient probe or a 5-mm
PFG AutoXID (1H/X15N-31P) probe, respectively. Standard pulse se-
quences were used. The structures of the studied steroids were de-
termined via analysis of homonuclear (DQF-COSY, zTOCSY, and
NOESY) and heteronuclear (1H−13C HSQC, HSQC-TOCSY, and HMBC)
spectra. The 1H connectivities were established via analysis of the DQF-
COSY and TOCSY spectra. The resonances of all carbons with directly
attached protons were assigned using the HSQC and HSQC-TOCSY
spectra. Then, the HMBC spectra were used to assign the quaternary
carbon resonances and to evaluate the correctness of the connectivities
that were established via the analysis of the other spectra. NOESY
spectra were also utilized to establish the α/β deuterations of steroids
10–15.
2.4.3. Preparative separation of 3β,3′β-disteryl ether
Preparative chromatography was conducted using
a LC-20
Prominence (Shimadzu) system, which consisted of a pump, an auto-
sampler (ambient temperature), a thermostat for the column (25 °C), a
diode array detector (DAD, 210 nm) and a fraction collector. 3β,3′β-
Disteryl ethers were separated on a silver-impregnated silica column,
namely, Zorbax Si of dimensions 250 × 4.6 mm 5 µm, using 0,20%
MTBE in n-Hep. The major fractions were collected into separate col-
lector vials. After several injections and runs, the organic solvents in the
collected fractions were removed by a gentle nitrogen stream with no
heat.
2.4.4. GC–MS method for 3β,3′β-disteryl ethers
A Shimadzu GC-2010 Plus with an Optic-4 autosampler that was
coupled with an MS-TQ8050 mass spectrometer was used. 3β,3′β-dis-
teryl ethers were separated on ZB-1HT of dimensions 30 m × 0.25 mm
with a film thickness of 0.1 μm. The carrier gas was helium with an inlet
pressure of 318 kPa and a flow rate of 68.6 cm/s. The initial column
temperature of 280 ˚C was held for 2 min and increased by 10 °C/min to
390 °C, where it was held for 10 min. Injections (20.0 µL) were con-
ducted in a split program mode (Table S7). The injector temperature
was 70 °C for 45 s, with the solvent valve open, which was increased
with a temperature gradient of 40 °C/s to 360 °C. The transfer line
temperature was 325 °C, the ion source temperature was 280 °C and the
energy of the electron impact ionization was 70 eV. The 3β,3′β-disteryl
ethers were analysed using selected ion monitoring (SIM). For 3β,3′β-
dicholesteryl ether, 369 m/z were monitored, for 3β,3′β-distigmasteryl
ether 394 m/z and for 3β,3′β-disitosteryl ether 397 m/z.
The experiments were conducted under the following conditions:
DQF-COSY - spectral widths of 5000 Hz in both dimensions, 4096
complex points in t2, 1024 complex points in t1, 1 scan per increment,
and a relaxation delay of 1 s; TOCSY - spectral widths of 5000 Hz in
both dimensions, 2048 complex points in t2, 1024 complex points in t1,
2 scans per increment, a relaxation delay of 1 s and a spin-lock time of
20 ms; NOESY - spectral widths of 5000 Hz in both dimensions, 1024
complex points in t2, 512 complex points in t1, 4 scans per increment, a
relaxation delay of 1 s and a mixing time of 400 ms; 1H–13C HSQC -
spectral widths of 5000 Hz in F2 and 19000 Hz in F1, 1024 complex
points in t2, 512 complex points in t1, 2 scans per increment, and a
relaxation delay of 1 s; 1H–13C HSQC-TOCSY - spectral widths of
5000 Hz in F2 and 15000 Hz in F1, 1024 complex points in both di-
mensions, 2 scans per increment, a relaxation delay of 1 s and a spin-
lock time of 18 ms; and 1H–13C HMBC - spectral widths of 5000 Hz in
F2 and 25000 Hz in F1, 1024 complex points in both dimensions, and 4
scans per increment.
2.4.5. Method validation
The method validation was described in previous works (Szterk,
Bus, et al., 2018; Szterk, Zmysłowski, et al., 2018). The linearity was
evaluated, and the limits of detection (LOD) and limits of quantitation
(LOQ) were calculated based on the standard deviation of the response
and the slope of the analytical curve. Recovery tests were conducted
using three concentration levels: low concentration (25 ng/mL),
medium concentration (100 ng/mL) and high concentration (250 ng/
mL)
All the spectra were referenced indirectly (Wishart et al., 1995).
When necessary, the data were processed using linear prediction in t1,
which was followed by zero-filling in both dimensions. Gaussian
weighting functions were applied in both domains prior to Fourier
transformation.
Prior to the analysis, 10 to 20 mg of the samples were dissolved in
600 µL of deuterated chloroform‑d1.
2.4.6. LC-QTOF-MS methods
2.6. Biological studies
A
UHPLC Ultimate 3000 (Dionex Thermo Fisher Scientific,
Sunnyvale, California, USA) system, which consisted of a pump, a de-
gasser, an autosampler, and a column heater instrument, was used.
Data processing was conducted with Chromeleon 6.8 and Chromeleon
Validation ICH software (Dionex). To determine the peak elution order,
2.6.1. Cell culture
Endothelial cells were grown in a medium that consisted of the
following for the in vitro culture congaing: EGF, FGF-2, VEGF, vitamin
C, R3-IGF-1, 3% FBS, gentamycin/amphotericin, hydrocortisone, and
heparin. The cells were kept in a 37 °C incubator with 95% humidity
and 5% CO2.
a
mass spectrometer maXis 4G from Bruker Daltonic (Billerica,
Massachusetts, USA) was used. The QTOF settings were as follows: at-
mospheric pressure chemical ionization (APCI) in positive ion mode,
nebulizer 2.0 bar, dry gas (nitrogen) flow rate of 4.0 L/min, dry heater
at 200 °C, vaporizer temperature 450 °C, capillary voltage of 3000 V,
corona discharge 3000nA and end plate offset of −500 V. The MS data
were recorded in full scan mode (from 100 to 1600 m/z). The mass
spectrometer was used in high-resolution mode (R = 60000), and an
internal calibrant (APCI/APPI calibrant) was used to obtain a precise
mass measurement. Chromatographic separation was conducted using a
Zorbax XBD-C18 50 × 2.1 mm with a 1.8 µm particle size in a gradient
system (phase A: MeOH, phase B: CH2Cl2). The following gradient
2.6.2. Liposome preparation
The DOPC: 3β,3′β-disteryl ether: DSG-PEG 2000 with a molar ratio
of 3:1:0.3 was dissolved in CHCl3, a Rotavapor was used to evaporate
the solvent, and the thin lipid film was rewetted in 10 mL of the culture
medium. The liposomes were prepared using ultrasound and an ex-
trusion method that utilized nylon membranes (10 µm, 0.45 µm, and
0.22 µm).
The control liposomes and the liposomes that contained 7-kCh (22)
were prepared via the same approach by exchanging the 3β,3′β-disteryl
4

A. Zmysłowski, et al.
FoodChemistry329(2020)127132
Table 2
Validation parameters for GC–MS analysis of the studied 3β,3′β-disteryl ethers.
Linearity range
Equation
R2 (n = 8)
0.9996
LOD [ng/mL]
3.7
LOQ [ng/mL]
12.3
Recovery
3β,3′β-dicholesteryl ether (dCh)
3β,3′β-distigmasteryl ether (dStig)
3β,3′β-disitosteryl ether (dSito)
15–250 ng/mL
y = 0.0188x + 0.1749
25 ng/mL
100 ng/mL
250 ng/mL
25 ng/mL
100 ng/mL
250 ng/mL
25 ng/mL
100 ng/mL
250 ng/mL
90.5% 7.6%
93.5% 3.4%
92.7% 1.4%
89.3% 10.8%
94.4% 4.1%
95.0% 1.9%
80.5% 9.8%
94.6% 2.9%
98.3% 2.0%
25–250 ng/mL
20–250 ng/mL
y = 0.0125x + 0.0158
y = 0.0192x + 0.1402
0.9981
0.9993
7.3
4.9
24.4
16.3
ether for cholesterol or 7-kCh (DOPC:sterol:DSG-PEG 2000) in a molar
ratio of 3:2:0.3.
3.2. Optimization of 3β,3′β-disteryl ether synthesis
To maximize the yield of the ether formation reaction, various
solvents and molar ratios of the substrate to the catalyst were com-
pared. The solvents that were investigated were aprotic solvents that
differed in terms of polarity: CH2Cl2, c-Hex, n-Hex, EtOAc, THF and
CHCl3. The temperature of the reaction was the boiling temperature of
the chosen solvent. The reaction was conducted until the substrate re-
acted, which was checked via TLC (n-Hex:EtOAc 3:1), or for 24 h. The
mass ratios of the substrate to the catalyst that were investigated, along
with the corresponding yields, are presented in Table S8. The yield of
each reaction was determined via HPLC analysis. Each time after the
reaction was stopped and catalyst was filtered off, a portion of 1 mL was
placed in a 10 mL glass vial, and the solvent of the reaction was eva-
porated. The residue was redissolved in 1 mL of n-Hep, and the samples
were injected for HPLC (Fig. S7). The yield of the reaction was calcu-
lated based on the peak area of the 3β,3′β-dicholesteryl ether that was
acquired after purification. Based on the yields of the reaction, the ratio
of the MK10 was chosen for the synthesis of all 3β,3′β-disteryl ethers.
CH2Cl2 produced the best results in terms of the yield of the reaction;
however, there were side products of similar polarity to 3β,3′β-dicho-
lesteryl ether, which rendered chromatographic purification more dif-
ficult. These products did not form with c-hex.
2.6.3. MTT test
The cell growth inhibition was determined via
a 3-(4,5-di-
methylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colori-
metric assay. Endothelial cells (HAEC) were plated in 96-well microtiter
plates with 8000 cells/well and allowed to attach for 24 h at 37 °C. The
test compounds in liposomes or dissolved in EtOH were added to the
cell culture medium to the desired final concentrations. After the cells
were cultured for 24 h, to the medium was added 10 μL of 5 mg/mL
MTT. After incubation for 4 h at 37 °C, 100 µL of 10% SDS in 0.01 N HCl
was added to the medium that contained MTT, and the plates were
shaken gently for an hour at room temperature. Absorbance values
were derived from the plate reading at 550 nm on a Biotek microtiter
plate reader (Winooski, VT). The reading reflected the number of viable
cells and was expressed as a percentage of the viable cells in the control.
3. Results
3.1. Sterol derivatives synthesis
As outlined in Fig. 1 the synthesis plan was divided into two main
routes. The first route was to synthesize the deuterium-labelled sterols
in order to prepare the 3β,3′β-d4-disteryl ethers. Conversion of sterols
to Δ4-3-ketosteroids (4, 5 and 6) using the Shvo’s catalyst in acetone as
hydrogen acceptor resulted in a simple and inexpensive alteration to
the Oppenauer oxidation with aluminum isopropoxide in acetone or
3.3. Method validation of isotope dilution gas chromatography mass
spectrometry
To analyse the 3β,3′β-disteryl ethers, gas chromatography-mass
spectrometry (GC–MS) with electron impact ionization and SIM was
used. The recoveries, linearity, LOD and LOQ were calculated. LOD,
LOQ, and recoveries on various levels and the linearity range of the
developed method are presented in Table 2. Chromatographs of the
3β,3′β-disteryl ethers are presented on Fig. 2. As was discussed, the
3β,3′β-disteryl ethers are difficult to analyse due to their instability in
various ion sources. Thus, gas chromatography must be employed with
high volume injection with a suitable split and temperature program in
the injector.
cyclohexanone (Almeida, Kočovský,
& Bäckvall, 1996; Ishihara,
Yamamoto, & Crich, 2009). In the next sequence of reaction, the enol
acetates (7, 8 and 9) were prepared by the reaction of the Δ4-3-ketos-
teroids with isopropenyl acetate followed by reduction with sodium
borodeuteride in deuterated ethanol (EtOD), which yielded 3,4-d2-
sterols (10, 11 and 12) (Gruenke & Craig, 1979). Reaction of obtained
3,4-d2-sterols in CH2Cl2 with MK10 catalyst resulted in 3β,3′β-d4-dis-
teryl ethers (13, 14, and 15).
The second route was to prepare the 7-ketosterols. Sterols were
protected as acetate (16, 17 and 18), which underwent an oxidation
reaction by tert-butyl hydroperoxide (t-ButOOH) with manganese (III)
acetate as a catalyst, to give the 7-keto derivatives (19, 20 and 21)
(Shing & Yeung, & Su, P. L. , 2006). The choice of the reaction was
based on the fewer side reactions due to mild oxidation, higher effi-
ciency and easier purification compared to the reactions using chro-
mium as an oxidizer (e.g., in the form of chromium oxide complex with
3.4. Analysis of a high-temperature treatment samples
Using prepared and validated method the 3β,3′β-disteryl ethers
were analysed in samples that were subjected to treatment at various
temperatures. The contents of each 3β,3′β-disteryl ether are presented
in Table 1. The highest contents of the dimers were identified in the
prepared samples that were treated for 1 h at 220 °C. These contents are
as follows: in cod-liver oil – 0.041 µg/g; butter – 0.374 µg/g; rapeseed
oil – 0.043 µg/g; and corn oil – 0.034 µg/g. The contents of the dimers
in the animal-origin samples are much higher than in the vegetable-
origin samples, and most of the contents of the prepared samples were
below LOD. The time-dependent 3β,3′β-disteryl formation is more
readily observable in Fig. 3, which presents the 3β,3′β-dicholesteryl
ether content in butter samples. The content at 180 °C is rising during
3,5-dimethylpyrazole or pyridinium chlorochromate) (Wendell
&
Edward, 2016). At this stage, a deprotection reaction in alkaline
medium led readily to 7-ketosterols (22, 23 and 24).
Additionally, epi-cholesterol (29) was obtained from cholesteryl
mesylate (28), by using KO2 as a nucleophile in the presence of 18-
crown-6 in the mixed solution of DMF and DMSO (You, Gong, & Sun,
2014).
5

A. Zmysłowski, et al.
FoodChemistry329(2020)127132
Fig. 2. Chromatograms of the studied 3β,3′β-disteryl ethers: (1) a chromatogram of the 3β,3′β-dicholesteryl ether standard; (2) a chromatogram of 3β,3′β-dis-
tigmasteryl ether; (3) a chromatogram of 3β,3′β-dististeryl ether; and (4) a chromatogram of the corn oil sample.
the time of heating from 0.116 µg/g to 0.200 µg/g, as also at 200 °C.
The same tendency is observed for the cod-liver oil. The measured
content at 220 °C peaks in 1 h of the treatment at 0.374 µg/g and
0.041 µg/g in butter and cod-liver oil, respectively. Additional heating
time resulted in significant reductions in the 3β,3′β-disteryl ether
content. The same is observed with the vegetable-origin samples;
however, after 2 h of heating, the measured contents were below LOD.
were dissolved in EtOH. The cytotoxic properties of 7-kCh are com-
parable between EtOH dissolution and incorporation in liposomes (Fig.
S2).
4. Discussion
4.1. Synthesis of the ethers and the ether formation mechanism
3.5. Effect of 3β,3′β-disteryl ethers on endothelial cells
It was reported previously that 3β,3′β-disteryl ethers were prepared
using reaction at high temperature with copper sulfate as a catalyst
(Schulte & Weber, 1987; Sosińska et al., 2013), which suggests that a
catalyst must be present to realize the formation of an ether bond (in
this case, the metal ion). The yield of the reaction was very low (< 5%),
and very tedious procedures were required for obtaining a pure com-
pound. 3β,3′β-dicholesteryl ether was also synthesized using direct
anodic oxidation of cholesterol in CH2Cl2 with 28% yield (Morzycki &
Sobkowiak, 2015). Despite this, only the use of the sterol reaction with
MK10 enabled the acquisition of 3β,3′β-disteryl ethers in pure form and
After incorporation into the bilayer liposomes, the synthesized
compounds were tested on endothelial (TeloHAEC) cells. The effects of
3β,3′β-disteryl ethers on the percent of live cells were compared to that
of the control, and they are presented in Fig. S1. Additionally, images
were taken before and after staining with propidium iodide to present
the changes in the morphology. To evaluate the suitability of the pre-
paration of the liposomes, the cytotoxic properties of the liposomes that
contained 7-kCh were compared to those of the same substance that
Fig. 3. Measured contents of 3β,3′β-dicholesteryl ether after treatment at various temperatures (180, 200 and 220 °C).
6

A. Zmysłowski, et al.
FoodChemistry329(2020)127132
in large quantity (Li, Li, Li, & Yang, 1998).
required in the biological matrix. Various metal ions (such as Al, Mg
and Fe) that are present in food could catalyse the elimination of the β-
hydroxyl moiety and the formation of an i-steroid intermediate, which
could create a bond with a partially negatively charged oxygen on
another sterol particle to form 3β,3′β-disteryl ether.
To realize the highest yield of the 3β,3′β-disteryl ethers formation
the various solvents were tested. The use of CH2Cl2 with MK10 pro-
duces the best results in terms of the yield of the reaction. The reaction
in c-hex has a lower yield, but the formation of side products is limited;
hence, the compounds are much easier to purify using preparative
chromatography.
Based on the conclusions that were obtained on the basis of the
reaction of sterols and stanols, namely, only one sterol molecule after
the dehydration reaction forms an i-steroid intermediate to form an
ether bond, an attempt was made to generate ether from the native
sterol (cholesterol) and oxidized sterol in carbon position 7 (7-ke-
tocholesterol, 22). Reaction with MK10 was conducted, and the ex-
pected product was obtained (3β,3′β-dicholesteryl ether), along with
oxidized 3β,3′β-disteryl ether (7-ketocholesteryl-cholesterol ether). The
structure was confirmed using NMR and LC-MS/MS (product 25).
Additionally, the oxidized 3β,3′β-disteryl ethers from phytosterol (β-
sitosterol and stigmasterol) and theirs oxidized analogues (7-ketosi-
tosterol and 7-ketostigmasterol, which correspond to 26 and 27, re-
spectively) were prepared. Compounds with such structure have been
identified using LC-MS/MS in previous studies; however, to the best of
our knowledge, they have not been obtained in pure form so far via
chemical synthesis (Rudzińska et al., 2009; Struijs et al., 2010). Since
one of the main reactions occurs during the high-temperature oxidation
of sterols, oxidized sterol dimers may form or the resulting 3β,3′β-dis-
teryl ether may be oxidized. Hence, oxidized 3β,3′β-disteryl ethers are a
new type of compounds that could originate from high-temperature
sterol treatment and could affect humans. Since sterols and oxysterols
originate from reactive oxygen species that are constantly associated
with atherosclerosis, oxidized 3β,3′β-disteryl ethers could have an im-
pact on atherosclerosis.
MK10 has been widely used in various reactions as a catalyst
(Kumar, Dhakshinamoorthy, & Pitchumani, 2014). MK10 is used as a
heterogeneous catalyst due to its favourable properties, such as low
cost, thermal stability, the presence of Lewis and Brønsted acid sites,
large surface area, and ease of separation; furthermore, it is en-
vironmentally benign. The reactions that are catalysed by MK10 are
typically conducted under mild conditions with high yields and high
selectivity, and the workups of these reactions are highly simple; only
filtration for removing the catalyst and evaporation of the solvent are
required. The structure of the catalyst is presented in Fig. S4. Based on
the composition of the catalyst, the metal ion that coordinates the ether
formation is likely aluminium. However, the use of aluminium chloride
(AlCl3) instead of MK10 in the reaction does not lead ether formation;
the only product of the reaction was the elimination product – cholesta-
3,5-diene. Moreover, the characteristic purple colour of the catalyst is
observed, as in the MK10 reaction (Photograph S1), which suggests that
aluminium or another metal ion in MK10 can be a catalyst in the ether
formation mechanism. We suggest that the Lewis site of the catalyst
with a metal ion is mainly responsible for the elimination of the β-
hydroxyl moiety and that the i-steroid intermediate is stabilized by a
tetrahedral sheet of Si(OH)4, which enables the oxygen from another
sterol molecule to attack the partially positively charged intermediate
to form an ether bond.
Due to the proven 3β,3′β-disteryl ether formation at high tem-
peratures, an attempt was made to analyse the reaction mechanism
using the prepared synthesis method. The reaction does not yield the
assumed product when only 5α-cholestan-3β-ol (cholestanol) is used.
However, a mixture of cholesterol and cholestanol yields 2 products –
3β,3′β-dicholesteryl ether and 3β,3′β-cholestenyl-cholesterol ether
(GC–MS chromatogram in Fig. S6). The ether was confirmed only via
the GC–MS method due to the inability to acquire dimers in pure form.
Based on that reaction, it was concluded that the double bond between
carbons 5 and 6 of the sterol moiety is essential for the ether formation.
Additionally, after synthesis of the epi-cholesterol, which was used as a
substrate for ether formation, again, no product was observed. Hence, it
is assumed that to form an ether, the double bond and a hydroxyl group
in position β must be present. The observed lack of reaction is likely
caused by the difficulty of leaving the hydroxyl group in the α position
(Sun, Cai, & Peterson, 2009). Additionally, to evaluate these assump-
tions, the dimeric products were prepared from a few additional sterol
derivatives with a double bond and a hydroxyl moiety in position β: 5-
pregnen-3β-ol-20-one, 3β,21-dihydroxy-5-pregnen-20-one 21-acetate,
16α,17α-epoxy-3β,21-dihydroxy-5-pregnen-20-one 21-acetate and
diosgenin (Table 3). From all substrates, it was possible to acquire
3β,3′β-disteryl ethers in pure form, with confirmation of the structure
via NMR. Diosgenin did not form the dimeric product in the reaction in
CH2Cl2; it only formed this product in c-hex. This is probably due to the
increased reaction temperature. Additionally, the low yield of 3β,3′β-
diosgenin ether could be due to low purity of the substrate, which might
contribute to side reactions.
4.2. Chromatography methods for analysing 3β,3′β-disteryl ethers
Due to the very low polarity of the 3β,3′β-disteryl ethers, a different
approach is required for analysis via chromatography and mass spec-
trometry. The typical ion source that is used for sterol analysis that is
coupled with liquid chromatography is APCI. However, chromato-
graphy of 3β,3′β-disteryl ether on C8 or C18 modified silica columns
requires the use of a high-elution-strength solvent (e.g., CH2Cl2 or
CHCl3), which unfortunately suppress the ionization of 3β,3′β-disteryl
ether, thereby resulting in a very poor LOD. The use of a shorter-chain
modified silica (e.g., C1 or C4) does not result in satisfactory resolution
between 3β,3′β-distigmasteryl and 3β,3′β-disitosteryl ether. The use of
normal phase chromatography that is coupled with APCI mass spec-
trometry yielded similar results regarding LOD and LOQ to those of
reversed-phase chromatography. Additionally, instability in various ion
sources of mass spectrometers render compounds of this type very
complicated to analyse. The other method for analysing dimeric pro-
ducts, which was used earlier, is liquid chromatography that is coupled
with mass spectrometry using silver ion coupled adducts (Struijs et al.,
2010). To form such adducts, a volatile silver salt must be used, such as
AgBF4. However, due to the occurrence of two isotopes of silver,
namely, 107Ag and 109Ag, in a ratio of almost 50%, the abundance of
molecular ions is lower by 50%; hence, this method is mostly useful for
qualitative research. Therefore, based on the previously reported GC-
FID method (Schulte & Weber, 1987), which demonstrated that the
analysis of 3β,3′β-disteryl ethers is possible with gas chromatography, a
method that uses this technique was developed. To search for 3β,3′β-
disteryl ether that was formed after high-temperature treatment, gas
chromatography must be coupled with mass spectrometry to realize
enhanced sensitivity and selectivity. Unfortunately, due to the in-
stability of the compounds in the electron impact (EI) ionization
method, the LOD and LOQ are not high. EI-ionized sterols provide
spectra that are typical for hydrocarbons, with many fragments but
with no apparent precursor that could be used for MS/MS analysis
(Szterk & Pakuła, 2016). In the case of 3β,3′β-disteryl ethers, no mo-
lecular ions are observed, which limits the analysis to SIM of
However, as discussed, a structural requirement is needed on only
one particle, which is subjected to a reaction. The double bond and
hydroxyl group at position β are probably required for the formation of
an i-steroid intermediate ion, which attacks a partially negatively
charged oxygen atom on another sterol particle (Fig. S5). Such an i-
steroid intermediate product was also observed in a reaction on me-
sylate and in a tosylate cholesterol reaction with a strong Lewis acid by
Sun et al. (Sun et al., 2009). This is why only one particle must be
converted into i-steroid to form the ether bond. This may also be
7

A. Zmysłowski, et al.
FoodChemistry329(2020)127132
Table 3
Various substrates, products, and yields for 3β,3′β-disteryl formation via MK10 reaction.
Substrate
Yield (%)
Product
–
Structure
Name
cholesten-5-en-3α-ol(epi-cholesterol)
no reaction
5α-cholestan-3β-ol(cholestanol)
no reaction
–
cholesterol+5-cholesten-3β-ol-7-one (22, 7-ketocholesterol)
β-sitosterol+5-sitosten-3β-ol-7-one (23, 7-ketositosterol)
67%
61%
25
26
stigmasterol+5-stigmasten-3β-ol-7-one (24, 7-ketostigmasterol)
64%
27
5-pregnen-3β-ol-20-one
74%
68%
30
31
3β,21-dihydroxy-5-pregnen-20-one 21-acetate
16α,17α-epoxy-3β,21-dihydroxy-5-pregnen-20-one 21-acetate
53%
35%
32
33
(3β,25R)-spirost-5-en-3-ol(diosgenin)
8

A. Zmysłowski, et al.
FoodChemistry329(2020)127132
fragmented molecules. The use of a different ionization source, such as
the positive or negative chemical ionization (PCI and NCI with methane
as the reagent gas), yielded a very low molecular ion concentration,
which resulted in very high LOQ. Further optimization of the ion source
parameter did not improve the ionization; thus, the use of a chemical
ionization source was discontinued. Moreover, there is no available
oxygen atom for derivatization reactions for the realization of higher
resolution and/or sensitivity, which is a common approach in analysing
sterol compounds with GC–MS (Pizzoferrato, Nicoli, & Lintas, 1993).
Additionally, the high boiling point of 3β,3′β-disteryl ethers requires
the use of suitable high-temperature-resistance components (septum,
liners, and ferrules) and a high-temperature GC column (ZB-1HT). Due
to the low LOD and LOQ of the 3β,3′β-disteryl ethers using split/
splitless analysis, it was concluded that high-volume injection is ne-
cessary for enhancing the detection performance. The most important
parameters, which needed to be optimised were: the choose of the liner,
the solvent of the samples and the starting temperature of the injector
program. The usage of sintered glass liner with a taper in combination
with n-Hep as a sample solvent (n-Hep has relatively high boiling point
compared to other nonpolar solvents like CH2Cl2 or n-Hex) and opti-
mised program temperature (Table S7) allowed to increase the volume
of the injection to 20 µL to enhance the LOD and LOQ, which made
possible to analyse the low amounts of the 3β,3′β-disteryl ethers in
prepared samples.
may inhibit the ether bond formation between sterols moieties. How-
ever, it is also possible that ether formation reaction occurs in-
dependently of unsaturated fatty acids oxidation. Based on the hy-
pothesis that to form 3β,3′β-disteryl ethers in food samples metal ions
need to be present to catalyse the reaction of ether bond formation, the
different concentration of metals in the selected samples could be an-
other reason for obtained such different results. Copper, which was
used in the salt form as a catalyst for the synthesis of 3β, 3′β-disteryl
ethers, may differ 10 times in concentration between the butter (up to
1.72 ppm) and vegetable oils (up to 0.18 ppm) (Garrido, Frías, Díaz, &
Hardisson, 1994; Meshref, Moselhy, & Hassan, 2014). Therefore, the
different composition of metal ions could implicate the disproportion of
3β,3′β-disteryl ethers content, measured in selected samples.
Various 3β,3′β-disteryl ethers, in addition to 3β,3′β-distigmasteryl
ether and 3β,3′β-disitosteryl ether, were also detected in corn oils
(Fig. 2). However, due to the lack of the pure standards of the 3β,3′β-
disteryl ethers and based on the results regarding the differences in the
responses between 3β,3′β-distigmasteryl ethers and 3β,3′β-disitosteryl
ether, the quantitative analysis could not be conducted. Based on the
spectrum that was acquired of the visible peaks and of this additional
peak in a standard solution of 3β,3′β-disitosteryl ether, which was
formed during the synthesis with a common impurity of sitosterol -
stigmasterol, it was assumed that those ethers were 3β,3′β-stigmasteryl-
sitosterol ether and 3β,3′β-campesteryl-sitosterol ether.
4.3. Analysis of high-temperature treatment
4.4. Effect of 3β,3′β-disteryl ethers on endothelial cells
As presented in Table 1, the content of 3β,3′β-disteryl ethers varies
according to the temperature and duration of the heat treatment of the
samples. Based on the presented results, there is a noticeable trend of
increasing content at 180 °C and 200 °C during heating. However, the
highest content of 3β,3′β-disteryl ethers was observed at 1 h with a
heating temperature of 220 °C. The prolonged heating resulted in de-
creased content, which can be correlated with side reactions of native
sterols or reactions of dimers. It was proved that at such high tem-
peratures, fragmentation of the sterol groups and/or further poly-
merization occurs (Derewiaka & Molińska (née Sosińska), 2015).
Derewiaka has studied cholesterol dimer formation, especially in
cholesterol standards. The highest total dimer content was observed
after 1 h of processing at 220 °C and reached 184.4 mg/g of the non-
heated standard, which is in accordance with our findings (Derewiaka
& Molińska (née Sosińska), 2015). Recently, it was proven that dimer
cholesterol can be formed in butter after heating to 180 °C (Derewiaka,
2019). However, the method that was used to analyse the dimer for-
mation was gel permeation chromatography (GPC), which separates the
analysed substances in the samples according to differences in mole-
cular mass. Unfortunately, the structures of the analysed substances
were not confirmed. Characterization of 3β,3′β-disteryl ethers via mass
spectrometry is challenging, and they can be easily mistaken for the
native sterol of sterol derivative molecules.
Due to the proven formation of 3β,3′β-disteryl ethers during high-
temperature treatment of sterols, a preliminary study was conducted to
investigate their impact on human cells. Sterols and sterol derivatives
(e.g., oxysterols) are often associated with atherosclerosis. To in-
vestigate the influence of 3β,3′β-disteryl ethers on the eventual initia-
tion of atherosclerosis, endothelial cells were selected based on their
participation in plague formation (Zmysłowski & Szterk, 2017). The cell
line that was used in the study was TeloHAEC, which is a clonal cell line
that was immortalized by stably expressing human telomerase catalytic
subunit hTERT. The selection of the cell line was based on proven in-
volvement of the endothelium in atherosclerosis lesion initiation and
progression. Due to the non-polarity of 3β,3′β-disteryl ethers, suitable
sample preparation must be conducted. To evaluate the preparation of
the liposomes, the oxidation product of cholesterol – 7-kCh – was in-
corporated into the liposomes, and the cytotoxic properties were
compared to those of the control and those of oxysterol that was dis-
solved in EtOH. The cytotoxic properties of 7-kCh were comparable to
those of HAEC cells that were either in liposomes or dissolved in EtOH
(Fig. S2.).
The cells were incubated for 72 h with tested compounds in pre-
pared liposomes; however, based on data that were acquired via the
MTT test, no changes in cell viability compared to the control were
observed. Additionally, the morphology of the cells was investigated.
Images that were acquired after incubation with 3β,3′β-disteryl ethers
and images of the control are presented in Fig. S3. Changes are ob-
served in the morphology of the cell compared to the control, which are
probably due to the incorporation of the ethers into the lipid bilayer,
which impedes the fluidity of the membrane. However, based on
staining with propidium iodide, the cells still maintain the continuity of
the cell membrane. It also should be pointed out that the concentration
of 3β,3′β-disteryl ethers tested on endothelial cells was significantly
higher than concentration measured in selected samples. At a con-
centration comparable to that determined in the selected samples, no
biological effect was found. Therefore, the concentration was increased
to observe if the 3β, 3′β-disteryl ethers have any cytotoxic properties.
However, based on the acquired data, it was concluded that 3β,3′β-
disteryl ethers under the selected experimental conditions does not
possess toxic properties to endothelial cells. It was already shown that
3β,3′β-dicholesteryl ether does not cause toxicity to rats after
Despite a higher concentration of phytosterols in oils per gram
compared to the concentration of cholesterol in butter, the measured
amounts of 3β,3′β-disteryl ethers are much lower. Different composi-
tions of fatty acids in fats of animal and vegetable origin in selected
samples may be one of the reasons for the obtained results. In animal
origin samples, the content of saturated fats predominates. The but-
terfat mainly consists of palmitic, myristic, and stearic acid with a low
concentration of unsaturated acids (Rutkowska & Adamska, 2011).
Cod-liver oil contains mainly stearic acid with the addition of un-
saturated acids like oleic, and eicosenoic acid (Pan, Ushio, & Ohshima,
2005). On the other hand, vegetable oils mainly consist of unsaturated
fatty acids, for rapeseed oil, it is oleic, linoleic, and linolenic acid
(Farahmandfar, Asnaashari, & Sayyad, 2015) and for corn oil, it is oleic
and linolenic acid (Hwang et al., 2011). The higher concentration of
unsaturated fatty acids in vegetable oils may cause double bond oxi-
dation to be the main reaction occurring thought heat treatment, which
9

A. Zmysłowski, et al.
FoodChemistry329(2020)127132
subcutaneous administration (Larsen & Barrett, 1944). However, this
report only discusses the effects of 3β,3′β-disteryl ethers on rats after
subcutaneous administration, which is not a typical route by which
compounds of this type enter the body.
2014.08.117.
Derewiaka, Dorota (2019). Formation of Cholesterol Oxidation Products, Cholesterol
Dimers and Cholestadienes After Thermal Processing of Cholesterol Standards and
Butter. European Journal of Lipid Science and Technology, 121(9), 1800373. https://
doi.org/10.1002/ejlt.201800373.
Derewiaka, Dorota, & Obiedziński, M. (2010). Cholesterol oxides content in selected
animal products determined by GC-MS. European Journal of Lipid Science and
Technology, 112(10), 1130–1137. https://doi.org/10.1002/ejlt.200900238.
Derina, K., Korotkova, E., Taishibekova, Y., Salkeeva, L., Kratochvil, B., & Barek, J.
(2018). Electrochemical nonenzymatic sensor for cholesterol determination in food.
Analytical and Bioanalytical Chemistry, 410(20), 5085–5092. https://doi.org/10.
1007/s00216-018-1164-x.
5. Conclusions
Using a reaction of sterols with the MK10 catalyst, various 3β,3′β-
disteryl ethers were prepared. Additionally, a reaction mechanism that
is based on various sterol analogues of ether bond formation was pro-
posed, which could explain how the 3β,3′β-disteryl ethers are formed in
high-temperature-treatment samples. The GC–MS method for analysing
3β,3′β-disteryl ethers was developed and validated. It was proven that
such compounds can be formed via the high-temperature treatment of
biological type samples. The concentration of ethers depended on the
duration and the temperature of the treatment. The most 3β,3′β-disteryl
ethers are produced at 220 °C after 1 h of heating; however, it is pos-
sible that formed dimers may oxidize and that the native dimer con-
centration is decreasing over time. Additionally, to the best of our
knowledge, this is the first report that describes the concentration of
3β,3′β-disteryl ethers (specific steryl dimers) after the thermal heating
of high-concentration sterol samples. This is because most publications
are focused on analysing sterol dimers, without structural confirmation
of the analysed compounds. The influence of 3β,3′β-disteryl ethers was
also evaluated with endothelium cells; however, for those cells, no
toxicity was observed on the tested substances.
Farahmandfar, R., Asnaashari, M., & Sayyad, R. (2015). Comparison antioxidant activity
of Tarom Mahali rice bran extracted from different extraction methods and its effect
on canola oil stabilization. Journal of Food Science and Technology, 52(10),
6385–6394. https://doi.org/10.1007/s13197-014-1702-2.
Garrido, M. D., Frías, I., Díaz, C., & Hardisson, A. (1994). Concentrations of metals in
vegetable edible oils. Food Chemistry, 50(3), 237–243. https://doi.org/10.1016/
0308-8146(94)90127-9.
Gruenke, L. D., & Craig, J. C. (1979). The synthesis of cholesterol-2,2,4,4,6–d5. Journal of
Labelled Compounds and Radiopharmaceuticals, 16(3), 495–500. https://doi.org/10.
1002/jlcr.2580160316.
Hwang, J., Chang, Y.-H., Park, J. H., Kim, S. Y., Chung, H., Shim, E., & Hwang, H. J.
(2011). Dietary saturated and monounsaturated fats protect against acute acet-
aminophen hepatotoxicity by altering fatty acid composition of liver microsomal
membrane in rats. Lipids in Health and Disease, 10(1), 184. https://doi.org/10.1186/
1476-511X-10-184.
Ishihara, K., Yamamoto, H., & Crich, D. (2009). Aluminum Isopropoxide. In Encyclopedia
of Reagents for Organic Synthesis. American Cancer Society. https://doi.org/10.
1002/047084289X.ra084.pub2.
Kulig, W., Cwiklik, L., Jurkiewicz, P., Rog, T., & Vattulainen, I. (2016). Cholesterol oxi-
dation products and their biological importance. Chemistry and Physics of Lipids, 199,
144–160. https://doi.org/10.1016/j.chemphyslip.2016.03.001.
Kumar, B. S., Dhakshinamoorthy, A., & Pitchumani, K. (2014). K10 montmorillonite clays
as environmentally benign catalysts for organic reactions. Catalysis Science &
Technology, 4(8), 2378–2396. https://doi.org/10.1039/C4CY00112E.
Lampi, A.-M., Kemmo, S., Mäkelä, A., Heikkinen, S., & Piironen, V. (2009). Distribution of
monomeric, dimeric and polymeric products of stigmasterol during thermo-oxida-
tion. European Journal of Lipid Science and Technology, 111(10), 1027–1034. https://
doi.org/10.1002/ejlt.200900069.
CRediT authorship contribution statement
Adam Zmysłowski: Conceptualization, Investigation, Validation,
Writing - original draft. Jerzy Sitkowski: Investigation. Katarzyna
Bus: Investigation. Karol Ofiara: Investigation. Arkadiusz Szterk:
Supervision.
Larsen, C. D., & Barrett, M. K. (1944). Administration of 3, 5-Cholestadiene and
Dicholesteryl Ether to Mice and Rats. JNCI: Journal of the National Cancer Institute,
4(6), 587–594. https://doi.org/10.1093/jnci/4.6.587.
Li, J.-T., Li, T.-S., Li, L.-J., & Yang, Z.-Q. (1998). Ultrasound-promoted preparation of
disteryl ethers catalyzed by montmorillonite K 10. Ultrasonics Sonochemistry, 5(2),
83–85. https://doi.org/10.1016/S1350-4177(98)00010-8.
Declaration of Competing Interest
Menéndez-Carreño, M., Ansorena, D., Astiasarán, I., Piironen, V., & Lampi, A.-M. (2010).
Determination of non-polar and mid-polar monomeric oxidation products of stig-
masterol during thermo-oxidation. Food Chemistry, 122(1), 277–284. https://doi.org/
10.1016/j.foodchem.2010.01.063.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Meshref, A. M. S., Moselhy, W. A., & Hassan, N. E.-H. Y. (2014). Heavy metals and trace
elements levels in milk and milk products. Journal of Food Measurement and
Characterization, 8(4), 381–388. https://doi.org/10.1007/s11694-014-9203-6.
Morzycki, J. W., & Sobkowiak, A. (2015). Electrochemical oxidation of cholesterol.
Beilstein Journal of Organic Chemistry, 11, 392–402. https://doi.org/10.3762/bjoc.
11.45.
Acknowledgement
The author is grateful to Prof. Sylwia Flis for helping with the cul-
turing of the HAEC cells.
Pan, X.-Q., Ushio, H., & Ohshima, T. (2005). Comparison of volatile compounds formed
by autoxidation and photosensitized oxidation of cod liver oil in emulsion systems.
Fisheries Science, 71(3), 639–647. https://doi.org/10.1111/j.1444-2906.2005.
01010.x.
Funding
Pedersen, D. S., & Rosenbohm, C. (2001). Dry Column Vacuum Chromatography.
Synthesis, 2001(16), 2431–2434. https://doi.org/10.1055/s-2001-18722.
Pizzoferrato, L., Nicoli, S., & Lintas, C. (1993). GC-MS characterization and quantification
of sterols and cholesterol oxidation products. Chromatographia, 35(5), 269–274.
https://doi.org/10.1007/BF02277508.
This study was supported by a grant for Young Scientists, namely,
grant number 1.34, and co-financially supported by Polish Ministry of
Science and Higher Education statutory grant number 9.52/19.
Rudzińska, M., Przybylski, R., & Wąsowicz, E. (2009). Products Formed During Thermo-
oxidative Degradation of Phytosterols. Journal of the American Oil Chemists’ Society,
86(7), 651–662. https://doi.org/10.1007/s11746-009-1397-0.
Appendix A. Supplementary data
Rutkowska, J., & Adamska, A. (2011). Fatty Acid Composition of Butter Originated from
North-Eastern Region of Poland. Polish Journal of Food and Nutrition Sciences, 61(3),
187–193. https://doi.org/10.2478/v10222-011-0020-x.
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.foodchem.2020.127132.
Schött, H.-F., & Lütjohann, D. (2015). Validation of an isotope dilution gas chromato-
graphy-mass spectrometry method for combined analysis of oxysterols and oxyphy-
tosterols in serum samples. Steroids, 99(Pt B), 139–150. https://doi.org/10.1016/j.
steroids.2015.02.006.
References
Schulte, E., & Weber, N. (1987). Analysis of disteryl ethers. Lipids, 22(12), 1049–1052.
https://doi.org/10.1007/BF02536449.
Almeida, M. L. S., Kočovský, P., & Bäckvall, J.-E. (1996). Ruthenium-Catalyzed
Oppenauer-Type Oxidation of 3β-Hydroxy Steroids. A Highly Efficient Entry into the
Steroidal Hormones with 4-En-3-one Functionality. The Journal of Organic Chemistry,
61(19), 6587–6590. https://doi.org/10.1021/jo960361q.
ShingYeung, T. K. M., & Su, P. L. (2006). Mild Manganese(III) acetate catalyzed allylic
oxidation: application to simple and complex alkenes. Organic Letters, 8(14),
3149–3151. https://doi.org/10.1021/ol0612298.
Sosińska, E., Przybylski, R., Aladedunye, F., & Hazendonk, P. (2014). Spectroscopic
characterisation of dimeric oxidation products of phytosterols. Food Chemistry, 151,
404–414. https://doi.org/10.1016/j.foodchem.2013.11.079.
Chien, J. T., Wang, H. C., & Chen, B. H. (1998). Kinetic Model of the Cholesterol
Oxidation during Heating. Journal of Agricultural and Food Chemistry, 46(7),
2572–2577. https://doi.org/10.1021/jf970788d.
Sosińska, E., Przybylski, R., Hazendonk, P., Zhao, Y. Y., & Curtis, J. M. (2013).
Characterisation of non-polar dimers formed during thermo-oxidative degradation of
β-sitosterol. Food Chemistry, 139(1–4), 464–474. https://doi.org/10.1016/j.
Derewiaka, D., & Molińska (née Sosińska), E. (2015). Cholesterol transformations during
heat treatment. Food Chemistry, 171, 233–240. https://doi.org/10.1016/j.foodchem.
10

A. Zmysłowski, et al.
FoodChemistry329(2020)127132
foodchem.2013.01.053.
Verleyen, T., Forcades, M., Verhe, R., Dewettinck, K., Huyghebaert, A., & Greyt, W. D.
(2002). Analysis of free and esterified sterols in vegetable oils. Journal of the American
Oil Chemists’ Society, 79(2), 117–122. https://doi.org/10.1007/s11746-002-0444-3.
Wendell, S., & Edward, J. (2016). A Short Review of Methods for the Allylic Oxidation of
Δ5 Steroidal Compounds to Enones. Journal of Steroids & Hormonal Science, 7(2),
https://doi.org/10.4172/2157-7536.1000171.
Struijs, K., Lampi, A.-M., Ollilainen, V., & Piironen, V. (2010). Dimer formation during the
thermo-oxidation of stigmasterol. European Food Research and Technology, 231(6),
853–863. https://doi.org/10.1007/s00217-010-1335-2.
Sun, Q., Cai, S., & Peterson, B. R. (2009). Practical Synthesis of 3β-Amino-5-Cholestene
and Related 3β-Halides Involving i-Steroid and Retro-i-Steroid Rearrangements.
Organic Letters, 11(3), 567–570. https://doi.org/10.1021/ol802343z.
Szterk, A., Bus, K., Zmysłowski, A., & Ofiara, K. (2018). Analysis of Menaquinone-7
Content and Impurities in Oil and Non-Oil Dietary Supplements. Molecules : A Journal
of Synthetic Chemistry and Natural Product. Chemistry, 23(5), https://doi.org/10.3390/
molecules23051056.
Wishart, D. S., Bigam, C. G., Yao, J., Abildgaard, F., Dyson, H. J., Oldfield, E., ... Sykes, B.
D. (1995). 1H, 13C and 15N chemical shift referencing in biomolecular NMR. Journal
of Biomolecular NMR, 6(2), 135–140. https://doi.org/10.1007/bf00211777.
You, Y. H., Gong, S. S., & Sun, Q. (2014). A Practical Synthesis of 3β-Amino-5-Cholestene.
Advanced Materials Research; Trans Tech Publications Ltd. doi: 10.4028/www.sci-
entific.net/AMR.830.151.
Szterk, A., & Pakuła, L. (2016). New method to determine free sterols/oxysterols in food
matrices using gas chromatography and ion trap mass spectrometry (GC-IT-MS).
Talanta, 152, 54–75. https://doi.org/10.1016/j.talanta.2016.01.049.
Szterk, A., Zmysłowski, A., & Bus, K. (2018). Identification of cis/trans isomers of me-
naquinone-7 in food as exemplified by dietary supplements. Food Chemistry, 243,
403–409. https://doi.org/10.1016/j.foodchem.2017.10.001.
Zmysłowski, A., & Szterk, A. (2017). Current knowledge on the mechanism of athero-
sclerosis and pro-atherosclerotic properties of oxysterols. Lipids in Health and Disease,
16(1), 188. https://doi.org/10.1186/s12944-017-0579-2.
Zmysłowski, A., & Szterk, A. (2019). Oxysterols as a biomarker in diseases. Clinica Chimica
Acta, 491, 103–113. https://doi.org/10.1016/j.cca.2019.01.022.
11