Food and Chemical Toxicology 83 (2015) 275e282

Contents lists available at ScienceDirect

Food and Chemical Toxicology

journal homepage: www.elsevier.com/locate/foodchemtox

Supercritical CO2 extraction of oil, fatty acids and flavonolignans from

milk thistle seeds: Evaluation of their antioxidant and cytotoxic
activities in Caco-2 cells
Naila Ben Rahal a, *, Francisco J. Barba b, Danielle Barth a, Isabelle Chevalot a
a
Laboratoire R
eactions et G
enie des Proc
ed
es UMR CNRS 7274, Universit e de Lorraine, 1, rue Grandville BP20451, 54001 Nancy, France
b
Nutrition and Food Science Area, Faculty of Pharmacy, Universitat de Val
encia, Avda. Vicent Andr
es Estell
es, s/n., 46100 Burjassot, Spain

a r t i c l e i n f o a b s t r a c t

Article history: The optimal conditions of supercritical carbon dioxide (SC-CO2) (160e220 bars, 40e80
C) technology
Received 26 May 2015 combined with co-solvent (ethanol), to recover oil, flavonolignans (silychristin, silydianin and silybinin)
Received in revised form and fatty acids from milk thistle seeds, to be used as food additives and/or nutraceuticals, were studied.
22 June 2015
Moreover, the antioxidant and cytotoxic activities of the SC-CO2 oil seeds extracts were evaluated in
Accepted 8 July 2015
Available online 11 July 2015
Caco-2 carcinoma cells. Pressure and temperature had a significant effect on oil and flavonolignans re-
covery, although there was not observed a clear trend. SC-CO2 with co-solvent extraction at 220 bars,
40
C was the optimum treatment to recover oil (30.8%) and flavonolignans from milk thistle seeds.
Keywords:
Milk thistle seeds
Moreover, linoleic (47.64e66.70%), and oleic (19.68e24.83%) acids were the predominant fatty acids in
Oil the oil extracts recovered from milk thistle under SC-CO2. In addition, SC-CO2 extract showed a high
Supercritical CO2 antioxidant activity determined by DPPH and ABTS tests. Cytotoxic activities of silychristin, silydianin
Fatty acids and silybinin and the obtained SC-CO2 extract (220 bars, 40
C) were evaluated against Caco-2 cells. The
Flavonolignans SC-CO2 extract inhibited the proliferation of Caco-2 cells in a dose-responsive manner and induced the
Caco-2 carcinoma cells highest percentage of mortality of Caco-2 cells (from 43 to 71% for concentrations from 10 up to 100 mg/
ml of SC-CO2 oil seeds)
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction in the anticholesterol diets for cardiovascular disease prevention

In the last two decades, the increasing ban on the use of artificial
food additives as well as the growing interest in use of active
compounds from natural sources that can contribute to promote
consumer's health and prevent diseases, has led to both food in-
dustry and food researchers to explore natural sources of food
additives and/or nutraceuticals (Bearth et al., 2014; Downham and
Collins, 2000; Parniakov et al., 2014; Thurmond, 2014).
For instance, Silybum marianum (family: Asteracae), commonly
known as milk thistle, is a native plant from the Mediterranean
area, which can constitute an important industrial agricultural
crop. Milk thistle leaves and flowers have been used as a vegetable
for salads and a substitute for spinach. On the other hand, S.
marianum seeds are roasted for use as a coffee substitute.
Moreover, the oil extracted from milk thistle seeds can be used
* Corresponding author.
E-mail address: nailabenrahal@gmail.com (N. Ben Rahal).
(El-Mallah et al., 2003) as it constitutes a good source of unsatu-
rated fatty acids (56% polyunsaturated and 21% monounsaturated)
(Yin et al., 1998), vitamin E (50e60 mg/100 g) (Hadolin et al., 2001),
phenolic compounds and flavonoids (0.25%) (Li et al., 2012; Pereira
et al., 2015; Wallace et al., 2003).
In addition, milk thistle seeds contain active compounds such as
silymarin. Silymarin has been known since centuries and recom-
mended in traditional European and Asian medicine, mainly for
treatment of liver disorders (Hadolin et al., 2001). These seeds
contain silymarin complex, which consists of four flavonolignans
(Pereira et al., 2015; Wallace et al., 2003): silychristin (SCN), sily-
dianin (SDN), silibinin (SBN) and taxifolin (TXF) Fig. 1.
The anticancer activity of silymarin as well as silibinin was
demonstrated against various cancer cells such as breast, skin, co-
lon, cervix, ovary, prostate, lung and hepatocellular cancers (Bosch-
Barrera and Menendez, 2015; Chu et al., 2004; Eo et al., 2015; Fan
et al., 2014; Jiang et al., 2015; Mastron et al., 2015; Sharma et al.,
2003; Thelen et al., 2004; Tyagi et al., 2004; Varghese et al., 2005).
Therefore, the recovery of these active compounds from milk

http://dx.doi.org/10.1016/j.fct.2015.07.006
0278-6915/© 2015 Elsevier Ltd. All rights reserved.
276 N. Ben Rahal et al. / Food and Chemical Toxicology 83 (2015) 275e282
Fig. 1. Flavonolignans structure (SBN: Silybinin, TXF: Taxifolin, SCN: Silychristin, SDN: Silydianin) (Quaglia et al., 1999).
thistle seeds constitutes an important challenge. Traditionally,
conventional solvent extraction has been used for these purposes
(Hadolin et al., 2001). However, conventional techniques involve
the use of toxic solvents (i.e. hexane) and long extraction times.
Thus, at this stage of development, there is a need to establish new
processes in full correspondence with green extraction concept,
which can reduce solvent consumption, extraction time and
toxicity (Chemat et al., 2012; Rosello-Soto et al., 2015).
In this way, supercritical fluid extraction (SFE) technique is a
promising alternative to conventional extraction methods (Rawson
et al., 2012). Carbon dioxide in its supercritical form (SCeCO2) has
been widely used in many SFE applications and, it has gained
increased attention in the food, pharmaceutical and cosmetic in-
dustries to obtain high-added value compounds such as flavors,
fragrance ingredients, food additives and/or nutraceuticals.
Some of the main advantages of using CO2 as supercritical fluid
are the low toxicity, and non-combustibility of this compound.
Moreover, CO2 presents an adjustable selectivity towards the
different components, and does not react with targeted compounds
(Friedrich and Pryde, 1984). Depending on pressure and tempera-
ture, liquid CO2 can be adjusted to be more liquid-like or gas-like
(Barba et al., 2014; Deng et al., 2014; Reverchon and Senatore,
1992). Moreover, the use of co-solvent gives an additional advan-
tage to SFE process as the solubility increase caused by co-solvent
allows pressure extraction decrease. This selectivity is due to spe-
cific interactions with the solute (Senorans et al., 2000).
To the best of our knowledge, there is a lack of information
about the effects of SFE with co-solvent on the recovery of fatty
acids, and flavonolignans from milk thistle seeds (Celik and Guru,
2015), as well as the potential cytotoxicity of these compounds
on cancer cell lines.
Thus, the aims of the present work are: 1) to evaluate the effects
of different SC-CO2 with co-solvent conditions to recover oil, fatty
acids and flavonolignans from S. marianum seeds. 2) To study the
antioxidant (DPPH and ABTS tests) and cytotoxic activities of S.
marianum oil seeds in Caco-2 carcinoma cells.
2. Materials and methods
2.1. Samples
Milk thistle samples were collected from the north of Tunisia.
Sheets, stems, roots and impurities were eliminated thanks to a
traditional sieve. Seeds were crushed to obtain a fine powder and
then stored at 4
C until needed for analysis. Seed powders were
separated into two sections according to their particle size (310 mm
and 620 mm of particle diameter). The values of dry matter (%) and
mineral content (%) of seed powders were 95.95 ± 0.36 and
4.81 ± 0.06, respectively.
2.2. SC-CO2 extraction
The experiments were carried out in a dynamic extraction unit
as it is shown in Fig. 2. The experimental design was previously
described (Bensebia et al., 2009). Thirty g of dried powder seeds
(particle diameter ¼ 310 mm) were packed into a sample unit. In
order to prevent the entrainment of milk thistle during extraction
process, and to improve the gas homogenization in the SC-CO2
equipment, glass wool was placed at the top and bottom of sample
unit. The sample was then allowed to reach the constant extraction
temperature before charging CO2 into the high-pressure pump
from the storage cylinder.
Ethanol (co-solvent) was added to increase polyphenols solu-
bility. In fact, SC-CO2 allows to solubilize nonpolar compounds (low
molecular weight) in low critical temperature (Tc ¼ 31
C), thus
facilitating the extraction of sensitive products at low temperature.
Ethanol was introduced with a Gilson pump with a flow rate of
5 ml/min for 15 min during contact time of the plant and SC-CO2
(before pumping the supercritical CO2).
The CO2 gas was further compressed to the desired pressure of
the pump. Static and dynamic times were 30 and 120 min (¼time to
exhaustion of oil in the seed), respectively. Samples were with-
drawn every 15 min. The separation of CO2 extract was carried out
at a temperature of 50
C. Co-solvent (ethanol) was removed using
a rotary evaporator. Finally, oil samples were kept in a freezer, until
needed for analysis. The operating conditions such as pressure,
temperature, density of solid and fluid, and solubility were fixed as
it is shown in Table 1.
2.3. Gas chromatographyemass spectroscopy (GCeMS) analysis
The GCeMS analysis was carried out on an Hewlett Packard
5890 II gas chromatograph with a HP-5MS 5% phenyl-
methylsiloxane capillary column (30 m  0.25 mm i.d., film thick-
ness 0.25 mm). Helium was used as a carrier gas at a flow rate of
20 mL/min. Each sample (1 mL) was injected into the column at a
split ratio of 15:1. The mass spectrometer was operated in electron-
impactionization (EI) mode by the energy of 70eV. The scanning
range was 50e600 amu and the scanning rate was 0.5s/scanning.
The individual identification of compounds was based on matching
 N. Ben Rahal et al. / Food and Chemical Toxicology 83 (2015) 275e282 277

Fig. 2. Dynamic extraction apparatus (Goodarznia and Eikani, 1998).

Table 1 2.5. Antioxidant tests

Molar volume of CO2 (MvCO2 ) and CO2 density (rCO2) under supercritical carbon
dioxide (SCeCO2) (160e220 bars, 40e80
C) conditions.
2.5.1. DPPH test
Pressure (bars) Temperature (
C) MvCO2 (cm3/mol) rCO2 (kg/m3) The evaluation of the antioxidant activity of silymarin was
160 40 60.43 730 performed with a DPPH radical and measured according to the
60 77.48 570 method described by Brand-Williams et al. (1995). The assays were
80 103.17 430 performed in 2 ml reaction mixtures containing 1.95 ml of 0.1 mM
180 40 57.83 760
DPPH-ethanol solution and 0.05 ml of the samples. The inhibitory
60 70.96 620
80 90.37 490 effect of the different concentrations of silymarin extracts
220 40 54.17 810 (0.05  102 mg/ml) on DPPH were measured by spec-
60 63.27 700 trophotometeric method. The absorbance of the reaction mixtures
80 75.83 580 at 517 nm was continuously monitored for 90 min.

2.5.2. ABTS test

their recorded mass spectra with those of NBS75K library data
provided by the software of GCeMS system (NIST, 1986).
2.4. High performance liquid chromatography (HPLC) analysis
HPLC method is adapted from Quaglia et al. (1999) method. A
Shimadzu LC-10AT VP chromatograph equipped with a Varian
photodiode array detector, was used for the analyses. The chro-
matograph was controlled and the data evaluated by a computer
Flyer Pentium, interface D 7000. Sample solutions were injected
using a 20 ml sample loop. The flavonolignans separation was car-
ried out using a stationary phase: C18 pre-column (Alltech) and C18
column (150 mm  4.6 mm, 5 mm) (Varian XRs) maintained at
40
C. The mobile phases used were described by Quaglia et al.
(1999). The elution was made at a flow rate of 1 ml/min in these
gradient conditions: Solvent A: Water, acidified with 10% H3PO4
(pH 2.6); Solvent B: acetonitril; Solvent C: methanol (flow rate of
1 ml/min). At t ¼ 0 min: 63% of solvent A, 15% of solvent B, 22% of
solvent C. At 7.5 min: 63% of solvent A, 15% of solvent B, 22% of
solvent C. At 8.5 min: 40% of solvent A, 20% of solvent B, 40% of
solvent C. At 15 min: 40% of solvent A, 20% of solvent B, 40% of
solvent C.
The photodiode array detector conditions were: l ¼ 330 nm.
Acquisition rate of spectra 1600 ms. Spectral bandwidth for each
channel 4 nm. Wavelength range: 220e350 nm. The standard so-
lutions were prepared from three flavonolignans (SBN 97.1%, SCN
82.2%, SDN 93.2%) (ChromaDex, France).
The standard TEAC (Trolox Equivalent Antioxidant Capacity)
assay described by Re et al. (1999) was used with minor mod-
ifications for the determination of TEAC value. The absorbance of
the mix solution (ABTS $þ radical þ phosphate buffer (pH 7.4)) was
measured at 734 nm (AABTS). One hundred microlitres of the diluted
sample were added in a tube, this solution was mixed quickly, and
the absorbance at 734 nm was measured after 60s. The decrease in
absorbance (DA ¼ AABTS  Asilymarin) after 60 s was calculated for
each diluted silymarin sample. The decrease in absorbance caused
by the addition of Trolox as the standard was measured by the same
procedure for each concentration of Trolox (50e400 mol/L) and the
calibration curve for the decrease in absorbance
(DA ¼ AABTS  ATrolox) of Trolox vs. Trolox concentration was con-
structed by linear regression. ABTSþ radical inhibition percentage
caused by each diluted sample was plotted against the non-diluted
sample volume in mixture reaction.
2.6. Cytotoxic activity
Caco-2 cells (colon cancer cell line) were cultivated in Dulbec-
co's modified eagle medium (DMEM) with high glucose (4.5 g/l)
(Sigma, Germany), and supplemented with 10% fetal calf serum
(FCS) (EuroBio, France), 2 mM L-glutamine, and 1% nonessential
amino acids (GIBCO, USA). The cells were usually splitted when
reaching confluence (5e7 days). They were first rinsed with Dul-
becco's phosphate-buffered saline without calcium (D-PBS) (Sigma,
Germany) and then trypsinised with a solution containing 0.25%
278 N. Ben Rahal et al. / Food and Chemical Toxicology 83 (2015) 275e282
trypsin and 1 mM EDTA (GIBCO, USA). Caco-2 cells were feeded into
96-well microplates at 1  104 cells/well in 200 ml of appropriate
culture medium.
The action of trypsin was stopped by the addition of culture
medium containing fetal calf serum (10%) and the cells were left in
solution. Subsequently, the cells were recovered by centrifugation
at 800 rpm/min for 5 min. Then, the cultures were maintained in an
incubator at 37
C under a moisture saturated atmosphere con-
taining 5% CO2. In order to ensure the nutrient supply to the cells
and removing cellular metabolites capable of inhibiting cell growth,
a change of culture medium was needed every 48 h. Cell growth
was evaluated daily using a microscope in opposite phase. The cells
were grown to 80% confluence where a concentration of
0.6  105 cells/cm was reached. After 24 h, the cells were exposed to
various concentrations of the compounds solubilized in DMSO
(final concentration of DMSO did not exceed 1%) and incubated for
48 h at 37
C, under 5% CO2 atmosphere.
Cell growth monitoring was determined by measurements in
microplates through a cellscreen® system. Cellscreen® consists of a
microscope coupled to an image analyzer. A camera allows taking
images of the surface of the wells of a microplate which is located
on a motorized stand. The images are analyzed well by well by the
processor and the surface of cell layer at the bottom of each well is
measured. Development cell is evaluated by image analysis by
determining the percentage of surface coated wells by cells.
On the other hand, the potential cytotoxicity of molecules
studied relative to Caco-2 cells in culture was determined using the
neutral red or NRU (Neutral Red Uptake) that can adsorb on lyso-
somal membrane of viable cells. The neutral red test was performed
after 48 h of incubation and the absorbance was measured at
540 nm. In case of death, the degraded lysosomal membrane does
not allow the adsorption of neutral red. The absorbance at 540 nm
is directly proportional to the amount of living cells (Mingoia et al.,
2007). The number of live cells after treatment with selected
Fig. 3. Oil recovery from milk thistle seeds using SC-CO2 process (160e220 bars, 40e80
C).
molecules to the number of untreated control concentration cells
allows the evaluation of the cytotoxicity of these molecules. The
results were expressed as IC50 mean values with the standard
deviations. IC50 was defined as the concentration of a molecule
leading to 50% cell mortality. The mortality rate (% cellular death) is
calculated according to the following equation:


. 
% cellular death ¼ 1  Absflavonolignans Absreference  100
where Absreference is the absorbance of control without fla-
vonolignans and Absflavonolignans is the absorbance obtained for the
assay with cells treated with flavonolignans).
2.7. Statistical analysis
All statistical analyses were performed using the software SPSS
Version 22 (IBM® SPSS® Statistics, USA). Significant differences
between the results were calculated by multiple sample compari-
son of the means (ANOVA) and the LSD test, with a significance
level of p < 0.05. The error bars presented on the figures correspond
to the standard deviations.
3. Results and discussion
3.1. Oil and fatty acids recovery
In order to establish the SC-CO2 optimum conditions to recover
oil from milk thistle seeds, different pressures (160e220 bars) and
temperatures (40e80
C) were studied (Table 1).
Two-way ANOVA analysis showed that pressure and extraction
temperature had a significant influence (p < 0.05) on the amount of
recovered oil (Fig. 3). As can be expected, when temperature was
40
C, oil recovery increased when pressure was augmented,
observing the maximum oil recovery (31.83 ± 0.84%) when SC-CO2
 N. Ben Rahal et al. / Food and Chemical Toxicology 83 (2015) 275e282 279

Table 2
Fatty acids recovery from milk thistle seeds using supercritical carbon dioxide (SCeCO2) process (160e220 bars, 40e80
C).

Pressure (bars) Temperature (
C) C16:0 C18:0 C18:1 C18:2 C20:0

160 40 8.74 ± 1.07abc 8.31 ± 1.97ab 22.56 ± 3.78ab 53.32 ± 8.59ab 2.41 ± 0.01a
60 10.04 ± 2.00ab 7.28 ± 2.53ab 19.68 ± 2.11a 59.94 ± 4.71ac 3.15 ± 0.71a
80 9.62 ± 0.46ab 6.15 ± 0.68ab 22.59 ± 0.65ab 47.64 ± 2.38b 2.87 ± 0.80a
180 40 9.12 ± 0.64abc 9.65 ± 4.71b 29.22 ± 4.71c 55.18 ± 9.95ab 3.08 ± 0.21a
60 7.44 ± 3.00bc 5.20 ± 0.42a 20.17 ± 0.42a 54.53 ± 3.00ab 2.31 ± 1.51a
80 8.72 ± 0.82abc 5.52 ± 0.59a 21.05 ± 1.01ab 53.54 ± 5.42ab 1.82 ± 0.55a
220 40 11.23 ± 0.56b 6.19 ± 1.23ab 21.61 ± 1.68ab 57.47 ± 3.20ac 2.33 ± 0.56a
60 6.46 ± 3.00c 9.43 ± 3.00b 19.77 ± 3.00a 66.70 ± 4.77c 2.83 ± 0.48a
80 10.89 ± 1.32b 8.11 ± 1.32ab 24.83 ± 1.32b 51.65 ± 1.60ab 2.17 ± 0.31a
aec
Different letters in the same column indicate significant statistical differences (p < 0.05).

Table 3
Flavonolignans recovery from milk thistle seeds using supercritical carbon dioxide (SCeCO2) process (160e220 bars, 40e80
C).

Pressure (bars) Temperature (
C) SCN (mg/ml of oil) SDN (mg/ml of oil) SBN (mg/ml of oil) SM (mg/ml of oil)
ab ab a
160 40 46.79 ± 5.63 107.03 ± 4.95 96.47 ± 4.80 253.63 ± 6.01ab
60 40.36 ± 1.58a 96.42 ± 4.70c 105.28 ± 5.73a 242.06 ± 2.23ac
80 48.31 ± 10.23abc 105.38 ± 6.20ab 70.20 ± 26.76b 223.90 ± 4.11cd
180 40 66.45 ± 3.00de 113.33 ± 3.00a 129.26 ± 3.00c 309.03 ± 9.01e
60 57.34 ± 8.57cdf 107.14 ± 6.82ab 105.88 ± 3.00a 270.36 ± 17.52bf
80 52.26 ± 8.15bc 94.29 ± 3.13c 74.88 ± 6.43b 221.43 ± 2.06cd
220 40 72.93 ± 4.71e 125.42 ± 4.71d 140.24 ± 4.99c 338.59 ± 6.12g
60 58.38 ± 2.99cdf 100.89 ± 4.46bc 131.49 ± 5.15c 290.76 ± 9.46ef
80 60.80 ± 1.32df 101.08 ± 1.32bc 51.10 ± 1.32d 212.98 ± 3.96d

SCN: Silychristin. SDN: Silydianin. SBN: Silybinin. SM: Silymarin.

aef
Different letters in the same column indicate significant statistical differences (p < 0.05).

at 220 bars and 40
C was used. under SC-CO2 conditions.

However, it should be noted that oil recovery was significantly
lower when temperature was increased for all pressures, especially
at 60
C. This fact can be explained by a decrease in density when
temperature was increased (Table 1), thus decreasing the oil solu-
bility. These results are in close agreement to those reported pre-
viously by Celik and Guru (2015) when they studied the impact of
SC-CO2 on oil recovery from milk thistle in similar conditions to
those evaluated in the present work. They found a significant
decrease in oil yield when temperature was increased. Moreover,
other authors also found a significant decrease in oil yield from
nutmeg seeds (Machmudah et al., 2006) and hemp seeds (Da Porto
et al., 2012) when temperature was increased from 40
C to 60
C
after applying 150 and 300 bars, respectively. So, it can be
concluded, that 40
C was the optimal temperature to recover oil
On the other hand, chromatographic analysis by GCeMS
revealed the presence of 7 fatty acids: myristic acid (C14:0), pal-
mitic (C16:0), stearic (C18:0), oleic (C18:1), linoleic (C18:2), lino-
lenic (C18:3), and arachidonic (C20:0) in all the SC-CO2 extracted
oils, being oleic and linoleic acids the predominant (Table 2).
Milk thistle oil presented an important content (70e85%) of
unsaturated fatty acids. These results are in close agreement to
those reported by Yin et al. (1998), who found values of unsaturated
fatty acids between 69 and 75% of total fatty acids and those found
by Li et al. (2012), who reported that unsaturated fatty acids were
around 74% in milk thistle oil seeds.
As can be seen in Table 2, linoleic acid is the most abundant acid
(47.64e66.70%) extracted in milk thistle seeds using SC-CO2. It
should be noted the importance of recovering this polyunsaturated
fatty acid as it is an essential fatty acid, which forms part of the
omega-3 family and has been involved in several physiological
functions (Denisa et al., 2013).
Oleic acid is also present at significant levels up to 19.68e24.83%
of SC-CO2 extracted oils. This fatty acid is also of a high importance
as can be used for both food and pharmaceutical industries. In this
line, essential fatty acids can provide nutritional and dietary values
and can justify its use for cardiovascular and drought and physio-
logical aging of the skin diseases (Calani et al., 2012). Moreover, the
results obtained in the present work showed that the extraction
temperature had no effect on the levels of SC-CO2 extractable fatty
acids.

3.1.1. Flavonolignans recovery

Fig. 4. HPLC analysis of flavonolignans from milk thistle oil seeds.
On the other hand, the optimum conditions to recover fla-
vonolignans from milk thistle under SC-CO2 extraction, were
evaluated. After applying SC-CO2, the HPLC analyses identified and
quantified the flavanolignan complex of silymarin (SM) (silychris-
tine (SCN), silydianin (SDN) and silybin (SBN)) (Table 3, Fig. 4.).
HPLC analysis showed that silychrisitin and silydianin levels
obtained in the present study are significantly higher (with
280 N. Ben Rahal et al. / Food and Chemical Toxicology 83 (2015) 275e282
Fig. 5. Antioxidant activity recovery from milk thistle seeds after applying supercritical
carbon dioxide (SCeCO2) with co-solvent (160e220 bars, 40e80
C).
32.31 ± 2.37 mg/g of seeds (72.93 ± 4.71 mg/ml oil) and
39.32 ± 1.31 mg/g of seeds (125.42 ± 4.71 mg/ml oil), respectively)
than those already reported by Quaglia et al. (1999) (0.47e6.9 mg/g
of seeds) and Subramaniam et al. (2008) (3.12 mg/g of seeds).
Silybin is the most active component of flavonolignans and its
quantification is a key step for the determination of biological ac-
tivity. Silybin is present at very high levels (40.98 ± 1.46 mg/g of
seeds (140.24 ± 4.99 mg/ml oil)) and is higher than those of Quaglia
et al. (1999) (1.43 mg/g of seeds) and Subramaniam et al. (2008)
(23.15 mg/g of seeds). This difference may be attributed to fla-
vonolignan's extraction method, pretreatment of seeds and seeds
origin (soil, climate, etc.).
As can be seen in Table 3, SBN and SDN were the predominant
flavonolignans extracted by SC-CO2 while the amount of SCN was
significantly lower. These results differ from those obtained from
Quaglia et al. (1999) who found that SBN was the predominant
flavonolignan found in milk thistle seeds followed by SDN and SCN.
These differences can be attributed to the extraction technique and
the origin of the plants used. In this line, other studies led to general
conclusions about the parameters according to the plant matrix
(oleoresins, essential, volatile and seed oils). The choice of extrac-
tion conditions depends on the specific compound. Thus the mo-
lecular weight and polarity should be taken into consideration
(Reverchon and De Marco, 2006).
Moreover, two-way ANOVA analysis was established to deter-
mine the effect of pressure and temperature on flavonolignans
recovery from milk thistle under SC-CO2. The results showed that
flavonolignans concentrations after SC-CO2 are greatly influenced
by the conditions of pressure and temperature, although the
behaviour varied according to the extracted flavonolignan (Table 3).
Fig. 6. Caco-2 cells specific growth rates (m, h1) in presence of flavonolignans and by
supercritical CO2 extract (SBN: Silybinin, SCN: Silychristin, SDN: Silydianin). E: Extract
obtained under SC-CO2 optimum conditions (220 bars, 40
C).
It is well established that supercritical fluid temperature to extract
thermolabile compounds must be fixed between 35
C and 60
C;
and near the critical point as low as possible to avoid degradation.
The increased temperature reduces SC-CO2 density (for a fixed
pressure) thereby reducing SC-CO2 solvent power; but it increases
vapor pressure of extracted compounds.
The results obtained in the present study were in close agree-
ment to those obtained by Celik and Guru (2015), when they
evaluated the effects of SC-CO2 (160e220 bars, 40e80
C) on the
recovery of silybin from milk thistle seeds. They also found a sig-
nificant decrease in the recovered amounts of silybin A and total
silybin when temperature was augmented up to 80
C compared to
lower temperatures. They attributed this phenomenon to a
decrease in CO2 density, thus difficulting the extraction of silybin.
3.2. Antioxidant activity
Milk thistle oil had significant scavenging effects on the DPPH
radical (data not shown). This antioxidant activity reflects the
bioactive molecules composition on oil extracted from S. marianum
(1.1 ± 0.6% of polyphenols on dried material, 13.4 mg b-carotene/
100 g of carotenoids and 70e85%% of unsaturated fatty acids).
However, the ANOVA analysis did not show any significant differ-
ences in DPPH values of the SC-CO2 oil extracts when pressure and
temperature were increased, thus obtaining DPPH values in the
same order of scavenging reactivity (0.57 ± 0.12 mg/ml).
On the other hand, as can be observed in Fig. 5, the ABTS values
were significantly higher when SC-CO2 treatments at 220 bars
(40
C and 60
C) were used compared to SC-CO2 at 160 and
180 bars. Moreover, the values at ABTS obtained after applying SC-
CO2 at 220 bars and 80
C were also lower than those obtained for
the same pressure conditions at lower temperatures. This fact can
be attributed to the degradation of some thermolabile and easily
oxidable compounds (i.e vitamin C, vitamin E) when high tem-
peratures and pressures conditions are combined.
3.3. Evaluation of cytotoxic activity
Moreover, the anti-proliferative activities of the individual fla-
vonolignans and the extracts obtained by SC-CO2 at the optimum
conditions (220 bars, 40
C) against Caco-2 cancer cells, were
evaluated (Fig. 6).
Results showed that silychristin and silydianin did not exert a
significant increase in antiproliferative properties when their con-
centrations were increased. However, silybin and milk thistle oil
Fig. 7. Mortality percentage of Caco-2 cells after 48 h of exposure with different
concentrations of flavonolignans determined by the neutral red test (SBN: Silybinin,
SCN: Silychristin, SDN: Silydianin). E: Extract obtained under SC-CO2 optimum con-
ditions (220 bars, 40
C).
 N. Ben Rahal et al. / Food and Chemical Toxicology 83 (2015) 275e282
seeds extract obtained after applying SC-CO2 at optimum condi-
tions showed a significant decrease in the proliferative activities in
the concentration range from 50 to 100 mg/ml. Silybin and SC-CO2
extract have cytostatic properties but silychristin and silydianin did
not exhibit any of these properties at the concentrations studied. In
this line, Hogan et al. (2007) also reported that silybin is the most
active flavonolignan of silymarin.
Moreover, in the present work, the effects of the individual
flavonolignans and the SC-CO2 extract obtained at 220 bar and
40
C and containing 32 mg/ml oil of silychristin, 39 mg/ml oil of
silydianin, and 46 mg/ml (oil) silybin on Caco-2 cancer cells growth
was observed after 48 h. In order to evaluate the cytotoxicity, the
percentage of mortality percentage was studied using neutral red
test. Fig. 7 presents the percentage of mortality of Caco-2 cancer
cells for different concentrations of flavonolignans standards and
SC-CO2 extract (E).
It was found that silybin, silychristin and SC-CO2 extract induced
a significant cell growth inhibition, thus leading to an arrest of cell
growth at different concentrations of silybin (10, 20, 50 and 100 mg/
ml), silychristin (50 and 100 mg/ml) and SC-CO2 extract (50 and
100 mg/ml) (Fig. 7). The apparent growth rate of the cells in the
presence of silychristin and silydianin was substantially similar to
the control cells showing that these two flavonolignans had no
significant effect on cells with mortality rates below 20%.
In terms of silybinin, which corresponds to the most active fla-
vonolignan, the cytotoxic effect at 50 and 100 mg/ml was significant
with 42% mortality and 52% mortality, respectively. The extract
obtained under SC-CO2 optimum conditions (220 bars, 40
C) also
gave reasonable recoveries of flavonolignans and inhibited the
proliferation of Caco-2 carcinoma cells in a dose-responsive
manner. As it is shown in Fig. 7, SC-CO2 extract induced an
important percentage of mortality (43e71 % for concentrations
from 10 to 100 mg/ml). This can be due to the content of fla-
vonolignans in the extract (silychristin 68 mg/ml, silydianin
121 mg/ml, and silybin 143 mg/ml).
Cell growth inhibition when SC-CO2 extract was used at low
concentrations can be explained by the fact that this extract is a
mixture of four flavonolignans, and perhaps a synergistic effect can
occur. Moreover, the existence of other active molecules in the
extract could explain the observed effects. The purpose of this
evaluation was to investigate the biological activity of the extract
obtained from milk thistle after applying SC-CO2 extraction process
with co-solvent at optimum conditions, which can have important
benefits from a healthy point of view.
Moreover, the IC50 values were calculated. It was found that oil
seed extract exerted stronger cytotoxicity than silibinin since the
IC50 values of cytotoxicity on Caco-2 were 96 mg/ml and 70 mg/ml,
for silibinin and oil seed extract, respectively. Thus, although sili-
binin showed its potential cytotoxicity, some other active com-
pounds may be responsible of the cytotoxicity of SC-CO2 extracts.
Our results were consistent with those obtained by Hogan et al.
(2007), which presented a cytotoxic effect of silybin on other co-
lon cancer cell lines Fet, Geo, and HCT116. Some other studies also
evaluated the cytotoxicity of silymarin towards different cancer
cells and determined the most effective concentration. In this line,
in a study performed on human epidermal A431 cells, it was
demonstrated that silymarin cytotoxicity varied according to the
concentration and exposure time (Ahmad et al., 1998). In addition,
the antiproliferative effect of silymarin on breast cancer cells MDA-
MB468 was demonstrated (Zi et al., 1998).
4. Conclusions
SC-CO2 with co-solvent technique is a useful tool to recover oil,
fatty acids and flavonolignans from milk thistle seeds. This
281
technique allowed the recovery of 7 fatty acids, being linoleic, oleic
and palmitic acids the predominant fatty acids identified in S.
marianum oil seeds. Moreover, four major flavonolignans (sily-
christin, silydianin, silybin and taxifolin) were recovered in the oil
extracts from milk thistle obtained under SC-CO2 extraction. Pres-
sure and temperature had a significant influence in the recovery of
the valuable compounds, reaching the maximum recovery of oil
and flavonolignans when SC-CO2 at 220 bars and 40
C was used.
Oil extracts obtained after SC-CO2 under the optimum conditions
presented an important antioxidant capacity. In addition, silybin
and SC-CO2 extracts showed important cytostatic properties on
Caco-2 cancer cell lines compared to silychristin and silydianin.
The results obtained in the present study are of a great interest
in the evaluation of the antioxidant and cytotoxic activities of fla-
vonolignans extracted by SC-CO2. Further experiments can be
considered towards other cancer cell lines to study the ability of
flavonolignans extracted by SC-CO2 to inhibit their proliferation.
Transparency document
Transparency document related to this article can be found
online at http://dx.doi.org/10.1016/j.fct.2015.07.006.
References
Ahmad, N., Gali, H., Javed, S., Agarwal, R., 1998. Skin cancer chemopreventive effects
of a flavonoid antioxidant silymarin are mediated via impairment of receptor
tyrosine kinase signaling and perturbation in cell cycle progression. Biochem.
Biophys. Res. Commun. 247, 294e301.
Barba, F.J., Grimi, N., Vorobiev, E., 2014. New approaches for the use of non-
conventional cell disruption technologies to extract potential food additives
and nutraceuticals from microalgae. Food Eng. Rev. 7, 45e62.
Bearth, A., Cousin, M.-E., Siegrist, M., 2014. The consumer's perception of artificial
food additives: influences on acceptance, risk and benefit perceptions. Food
Qual. Prefer 38, 14e23.
Bensebia, O., Barth, D., Bensebia, B., Dahmani, A., 2009. Supercritical CO2 extraction
of rosemary: effect of extraction parameters and modelling. J. Supercrit. Fluids
49, 161e166.
Bosch-Barrera, J., Menendez, J.A., 2015. Silibinin and STAT3: a natural way of tar-
geting transcription factors for cancer therapy. Cancer Treat. Rev. http://
dx.doi.org/10.1016/j.ctrv.2015.04.008.
Brand-Williams, W., Cuvelier, M.E., Berset, C., 1995. Use of a free radical method to
evaluate antioxidant activity. LWT Food Sci. Technol. 28, 25e30.
Calani, L., Brighenti, F., Bruni, R., Del Rio, D., 2012. Absorption and metabolism of
milk thistle flavanolignans in humans. Phytomedicine 20, 40e46.
Celik, H.T., Guru, M., 2015. Extraction of oil and silybin compounds from milk thistle
seeds using supercritical carbon dioxide. J. Supercrit. Fluids 100, 105e109.
Chemat, F., Vian, M.A., Cravotto, G., 2012. Green extraction of natural products:
concept and principles. Int. J. Mol. Sci. 13 (7), 8615e8627.
Chu, S.-C., Chiou, H.-L., Chen, P.-H., Yang, S.-F., Hsieh, Y.-S., 2004. Silibinin inhibits
the invasion of human lung cancer cells via decreased productions of
urokinase-plasminogen activator and matrix metalloproteinase-2. Mol. Carci-
nog. 40, 143e149.
Da Porto, C., Voinovich, D., Decorti, D., Natolino, A., 2012. Response surface opti-
mization of hemp seed (Cannabis sativa L.) oil yield and oxidation stability by
supercritical carbon dioxide extraction. J. Supercrit. Fluids 68, 45e51.
Deng, Q., Zinoviadou, K.G., Galanakis, C.M., Orlien, V., Grimi, N., Vorobiev, E.,
Lebovka, N., Barba, F.J., 2014. The effects of conventional and non-conventional
processing on glucosinolates and its derived forms, isothiocyanates: extraction,
degradation, and applications. Food Eng. Rev. 1e25. http://dx.doi.org/10.1007/
s12393-014-9104-9.
Denisa, I., Potierb, B., Vancassela, S., Heberdena, C., Lavialle, M., 2013. Omega-3 fatty
acids and brain resistance to ageing and stress: body of evidence and possible
mechanisms. Ageing Res. Rev. 12, 579e594.
Downham, A., Collins, P., 2000. Colouring our foods in the last and next millennium.
Int. J. Food Sci. Technol. 35, 5e22.
El-Mallah, M.H., El-Shami, S.M., Hassanein, M.M., 2003. Detailed studies on some
lipids of Silybum marianum (L.) seed oil. Grasas Aceites 54, 397e402.
Eo, H.J., Park, G.H., Song, H.M., Lee, J.W., Kim, M.K., Lee, M.H., Lee, J.R., Koo, J.S.,
Jeong, J.B., 2015. Silymarin induces cyclin D1 proteasomal degradation via its
phosphorylation of threonine-286 in human colorectal cancer cells. Int.
Immunopharmacol. 24, 1e6.
Fan, L., Ma, Y., Liu, Y., Zheng, D., Huang, G., 2014. Silymarin induces cell cycle arrest
and apoptosis in ovarian cancer cells. Eur. J. Pharmacol. 743, 79e88.
Friedrich, J.P., Pryde, E.H., 1984. Supercritical CO2 extraction of lipid-bearing ma-
terials and characterization of the products. J. Am. Oil Chem. Soc. 61, 223e228.
Goodarznia, I., Eikani, M.H., 1998. Supercritical carbon dioxide extraction of
282 N. Ben Rahal et al. / Food and Chemical Toxicology 83 (2015) 275e282
essential oils: modeling and simulation. Chem. Eng. Sci. 53, 1387e1395.
Hadolin, M., Skerget, M., Knez, Z., Bauman, D., 2001. High pressure extraction of
vitamin E-rich oil from Silybum marianum. Food Chem. 74, 355e364.
Hogan, F.S., Krishnegowda, N.K., Mikhailova, M., Kahlenberg, M.S., 2007. Flavonoid,
silibinin, inhibits proliferation and promotes cell-cycle arrest of human colon
cancer. J. Surg. Res. 143, 58e65.
Jiang, K., Wang, W., Jin, X., Wang, Z., Ji, Z., Meng, G., 2015. Silibinin, a natural
flavonoid, induces autophagy via ROS-dependent mitochondrial dysfunction
and loss of ATP involving BNIP3 in human MCF7 breast cancer cells. Oncol. Rep.
33, 2711e2718.
Li, F., Yang, L., Zhao, T., Zhao, J., Zou, Y., Zou, Y., Wu, X., 2012. Optimization of
enzymatic pretreatment for n-hexane extraction of oil from Silybum marianum
seeds using response surface methodology. Food Bioprod. Process 90, 87e94.
Machmudah, S., Sulaswatty, A., Sasaki, M., Goto, M., Hirose, T., 2006. Supercritical
CO2 extraction of nutmeg oil: experiments and modeling. J. Supercrit. Fluids 39,
30e39.
Mastron, J.K., Siveen, K.S., Sethi, G., Bishayee, A., 2015. Silymarin and hepatocellular
carcinoma: a systematic, comprehensive, and critical review. Anticancer. Drugs.
http://dx.doi.org/10.1097/CAD.0000000000000211.
Mingoia, R.T., Nabb, D.L., Yang, C.-H., Han, X., 2007. Primary culture of rat hepato-
cytes in 96-well plates: effects of extracellular matrix configuration on cyto-
chrome {P450} enzyme activity and inducibility, and its application in in vitro
cytotoxicity screening. Toxicol. Vitr. 21, 165e173.
NIST, 1986. National Bureau of Standards. National Institute of Standards and
Technologies (NIST), United States of America.
Parniakov, O., Barba, F.J., Grimi, N., Marchal, L., Jubeau, S., Lebovka, N., Vorobiev, E.,
2014. Pulsed electric field assisted extraction of nutritionally valuable com-
pounds from microalgae Nannochloropsis spp. using the binary mixture of
organic solvents and water. Innov. Food Sci. Emerg. Technol. 27, 79e85.
Pereira, C., Barros, L., Carvalho, A.M., Santos-Buelga, C., Ferreira, I.C.F.R., 2015. In-
fusions of artichoke and milk thistle represent a good source of phenolic acids
and flavonoids. Food Funct. 6, 56e62.
Quaglia, M.G., Bossu, E., Donati, E., Mazzanti, G., Brandt, A., 1999. Determination of
silymarine in the extract from the dried Silybum marianum fruits by high per-
formance liquid chromatography and capillary electrophoresis. J. Pharm. Bio-
med. Anal. 19, 435e442.
Rawson, A., Tiwari, B.K., Brunton, N., Brennan, C., Cullen, P.J., O'Donnell, C.P., 2012.
Application of supercritical carbon dioxide to fruit and vegetables: extraction,
processing, and preservation. Food Rev. Int. 28, 253e276.
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 Radic. Biol. Med. 26, 1231e1237.
Reverchon, E., De Marco, I., 2006. Supercritical fluid extraction and fractionation of
natural matter. J. Supercrit. Fluids 38, 146e166.
Reverchon, E., Senatore, F., 1992. Isolation of rosemary oil: comparison between
hydrodistillation and supercritical CO2 extraction. Flavour Fragr. J. 7, 227e230.
Rosello-Soto, E., Galanakis, C.M., Brncic, M., Orlien, V., Trujillo, F.J., Mawson, R.,
Knoerzer, K., Tiwari, B.K., Barba, F.J., 2015. Clean recovery of antioxidant com-
pounds from plant foods, by-products and algae assisted by ultrasounds pro-
cessing. Modeling approaches to optimize processing conditions. Trends Food
Sci. Technol. 42, 134e149.
Senorans, F.J., Ibanez, E., Cavero, S., Tabera, J., Reglero, G., 2000. Liquid
chromatographic-mass spectrometric analysis of supercritical-fluid extracts of
rosemary plants. J. Chromatogr. A 870, 491e499.
Sharma, G., Singh, R.P., Chan, D.C.F., Agarwal, R., 2003. Silibinin induces growth
inhibition and apoptotic cell death in human lung carcinoma cells. Anticancer
Res. 23, 2649e2655.
Subramaniam, S., Vaughn, K., Carrier, D.J., Clausen, E.C., 2008. Pretreatment of milk
thistle seed to increase the silymarin yield: an alternative to petroleum ether
defatting. Bioresour. Technol. 99, 2501e2506.
Thelen, P., Jarry, H., Ringert, R.-H., Wuttke, W., 2004. Silibinin down-regulates
prostate epithelium-derived ets transcription factor in LNCaP prostate cancer
cells. Planta Med. 70, 397e400.
Thurmond, T.S., 2014. U. S. Food and Drug Administration's (FDA) safety assessment
of food ingredients. In: ACS Symp. Ser, vol. 1162, pp. 91e95.
Tyagi, A.K., Agarwal, C., Chan, D.C., Agarwal, R., 2004. Synergistic anti-cancer effects
of silibinin with conventional cytotoxic agents doxorubicin, cisplatin and car-
boplatin against human breast carcinoma MCF-7 and MDA-MB468 cells. Oncol.
Rep. 11, 493e499.
Varghese, L., Agarwal, C., Tyagi, A., Singh, R.P., Agarwal, R., 2005. Silibinin efficacy
against human hepatocellular carcinoma. Clin. Cancer Res. 11, 8441e8448.
Wallace, S., Carrier, D.J., Beitle, R.R., Clausen, E.C., Griffis, C.L., 2003. HPLC-UV and LC-
MS-MS characterization of silymarin in milk thistle seeds and corresponding
products. J. Nutraceut. Funct. Med. Foods 4, 37e48.
Yin, Q., Wang, S., Nan, J., L.S, 1998. The fatty acid compositions of Silybum marianum
by GC-MS. J. Yanbian Univ. Sci. 24, 26e28.
Zi, X., Feyes, D.K., Agarwal, R., 1998. Anticarcinogenic effect of a flavonoid antioxi-
dant, silymarin, in human breast cancer cells MDA-MB 468: induction of G1
arrest through an increase in Cip1/p21 concomitant with a decrease in kinase
activity of cyclin dependent kinases and associated cyclins. Clin. Cancer Res. 4,
1055e1064.