Development and validation of an efficient ultrasound assisted extraction of phenolic compounds from flax (Linum usitatissimum L.) seeds

Ultrasonics Sonochemistry 26 (2015) 176–185

Contents lists available at ScienceDirect

Ultrasonics Sonochemistry journal homepage: www.elsevier.com/locate/ultson

Development and validation of an efficient ultrasound assisted extraction of phenolic compounds from flax (Linum usitatissimum L.) seeds Cyrielle Corbin a, Thibaud Fidel a, Emilie A. Leclerc a, Esmatullah Barakzoy a, Nadine Sagot b, Annie Falguiéres d, Sullivan Renouard a, Jean-Philippe Blondeau b, Clotilde Ferroud d, Joël Doussot a,c, Eric Lainé a, Christophe Hano a,⇑ a

Laboratoire de Biologie des Ligneux et des Grandes Cultures (LBLGC), UPRES EA 1207, Université d’Orléans, Chartres, France Conditions Extrêmes et Matériaux: Haute Température et Irradiation (CEMHTI) UPR3079, CNRS Orléans, France c Ecole Sciences industrielles et technologies de l’information (SITI), Département Chimie Alimentation Santé Environnement Risque (CASER), Le CNAM Paris, France d Service de Transformations Chimiques et Pharmaceutiques, ERL CNRS 3193, Le CNAM Paris, France b

a r t i c l e

i n f o

Article history: Received 19 November 2014 Received in revised form 17 February 2015 Accepted 24 February 2015 Available online 28 February 2015 Keywords: Ultrasound-assisted extraction Flaxseed Phenolic compounds Lignan Flavanol

a b s t r a c t Flaxseed accumulates in its seedcoat a macromolecular complex composed of lignan (secoisolariciresinol diglucoside, SDG), flavonol (herbacetin diglucoside, HDG) and hydroxycinnamic acids (p-couramic, caffeic and ferulic acid glucosides). Their antioxidant and/or cancer chemopreventive properties support their interest in human health and therefore, the demand for their extraction. In the present study, ultrasound-assisted extraction (UAE) of flaxseed phenolic compounds was investigated. Scanning Electron Microscopy imaging and histochemical analysis revealed the deep alteration of the seedcoat ultrastructure and the release of the mucilage following ultrasound treatment. Therefore, this method was found to be very efficient for the reduction of mucilage entrapment of flaxseed phenolics. The optimal conditions for UAE phenolic compounds extraction from flaxseeds were found to be: water as solvent supplemented with 0.2 N of sodium hydroxide for alkaline hydrolysis of the SDG–HMG complex, an extraction time of 60 min at a temperature of 25 °C and an ultrasound frequency of 30 kHz. Under these optimized and validated conditions, highest yields of SDG, HDG and hydroxycinnamic acid glucosides were detected in comparison to other published methods. Therefore, the procedure presented herein is a valuable method for efficient extraction and quantification of the main flaxseed phenolics. Moreover, this UAE is of particular interest within the context of green chemistry in terms of reducing energy consumption and valuation of flaxseed cakes as by-products resulting from the production of flax oil. Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction Flax (Linum usitatissimum L., Linaceae) is a common oilseed crop which seed are considered as a functional food [1]. Indeed, flaxseed Abbreviations: CafG, caffeic acid glucoside; DW, dry weight; FerG, ferulic acid glucoside; HCAG, hydroxy cinnamic acid glucosides; HDG, herbacetin di-glucoside; HMG, hydroxy methyl glutaryl; HMGA, hydroxy methyl glutaric acid; IS, internal standard; LOD, limit of detection; LOQ, limit of quantification; p-CouG, para-coumaric acid glucoside; RSD, relative standard deviation; SDG, secoisolariciresinol di-glucoside; SEM, Scanning Electron Microscopy; UAE, ultrasound-assisted extraction. ⇑ Corresponding author at: Laboratoire de Biologie des Ligneux et des Grandes Cultures (LBLGC), UPRES EA 1207, Université d’Orléans, Pôle Universitaire d’Eure et Loir, 21 rue de Loigny la Bataille, F-28000 Chartres, France. Tel.: +33 2 37 30 97 53; fax: +33 2 37 91 08 63. E-mail address: [email protected] (C. Hano). http://dx.doi.org/10.1016/j.ultsonch.2015.02.008 1350-4177/Ó 2015 Elsevier B.V. All rights reserved.

hulls represent a rich source of valuable phytochemicals such as lignan, flavonol and hydroxycinnamic acids [1,2]. Nowadays the preventive effects of flaxseed components on the onset of human diseases, especially the phytoestrogenic secoisolariciresinol diglucoside (SDG, Fig. 1a (1)), are well recognized and, for some, confirmed by both in vitro and in vivo studies [3]. Herbacetin diglucoside (HDG, Fig. 1a (2)) belongs to the flavonols and displays a large spectrum of biological activities, being the most active compound within the flavonoid group known for their cardiovascular beneficial effects [4]. The hydroxycinnamic acids stored as a glucoside form in flaxseeds, including p-coumaric (p-CouG, Fig. 1a (3)), caffeic (CafG, Fig. 1a (4)) and ferulic (FerG, Fig. 1a (5)) acids possess strong antioxidant properties [5] and could be very attractive in dermatology [6]. For all these properties, there has been an increasing concern in human consumption of flaxseed in the diet in

177

C. Corbin et al. / Ultrasonics Sonochemistry 26 (2015) 176–185

a

MeO

O O

HO

R R R

OH

O O

HO

O

OMe OH

OH (1)

O (2)

COOH

COOH

COOH

OH O

R

O

R

(3)

O

R

(4) O CH2

CH2 HO

R

(5)

OH

O

OMe

CH3

OH

(6)

b

c

Fig. 1. Structure of phenolic compounds involved in the lignan macromolecular complex. (a) Semi-developed structure of the complex components: (1) secoisolariciresinol di-glucoside (SDG), (2) herbacetin di-glucoside (HDG), (3) p-coumaric acid glucoside (pCouG), (4) cafeic acid glucoside (CafG), (5) ferrulic acid glucoside (FerG), (6) hydroxymethyl-glutaric acid (HMGA) (b) Schematic representation of lignan macromolecular complex described by Struis et al. (2008) (c) Representative chromatogram of a complete HPLC analysis after UAE of phenolics from flaxseed hulls where the main phenolic compounds considered in this study are indicated, respectively following their retention time: CafG, FerG, p-CouG, SDG and HDG. The o-coumaric acid was used as an internal standard (is).

order to improve the nutritional and health status [1]. Thus, there is a need for efficient extracting procedures of these phenolic compounds from seeds using environmental friendly solvents but also demanding less energy consuming.

In flaxseeds, these glucosylated phenolics are further stored as a hydroxymethyl glutaryl (HMG, Fig. 1a (6)) ester-linked complex named SDG–HMG (Fig. 1b). Extraction from this complex is traditionally achieved sequentially or simultaneously [7] by alcoholic

178

C. Corbin et al. / Ultrasonics Sonochemistry 26 (2015) 176–185

solid–liquid extraction and alkaline treatment. Several studies have been performed to optimize extraction and analysis of these valuable flax compounds using conventional methods [7–9], microwave-assisted extraction [2,4,10] or enzymatic-assisted extraction [11–13]. Methanol is traditionally used for the extraction of these compounds but due to its toxicity, other solvents are needed for food and cosmetic applications. Recent studies have evidenced that ultrasound application can enhance the extraction efficiency through acoustic cavitation and mechanical effects [14,15]. Moreover, the preservation of the efficacy while using green chemistry compatible solvent is possible with UAE. Acoustic cavitations produced by ultrasound can disrupt cell walls facilitating solvent penetration into the plant material and allowing the intracellular content to be released. Another mechanical effect caused by ultrasound could also be the agitation of the solvent used for extraction, thus increasing the surface contact area between the solvent and the targeted compounds by allowing greater penetration of the solvent into the sample matrix. Therefore the main advantages of ultrasound-assisted extraction (UAE) include reduced extraction time and reduced solvent consumption. In addition, UAE can be carried out at lower temperature which can avoid thermal damages to the extract and minimize the loss of bioactive compounds. In the present study, we report the optimization of UAE of SDG, HDG and hydroxycinnamic acid glucosides from flax hull quantified by HPLC with photodiode array detection and LC–ESI–MS. Various parameters influencing the extraction procedure such as solvent type, extraction time, ultrasound power and temperature as well as NaOH concentration for the release of phenolic compounds from the ester-linked complex were optimized. Finally, the resulting optimized UAE method was compared to traditional extraction methods in order to evaluate the enhancement of both protocol and yield. 2. Materials and methods 2.1. Plant material Flax (L. usitatissimum L.) seeds, cultivar Barbara, were provided by Coopérative Linière Terre de Lin (France). 2.2. Chemicals All reagents for extraction and HPLC analysis were of analytical grade or higher available purity and were purchased from Thermo Fisher Scientific. Deionized water was purified by a Milli-Q water-purification system from Millipore. All solutions prepared for HPLC were filtered through 0.45 lm nylon syringe membranes prior to use. All standards were purchased from Chromadex. 2.3. Apparatus Ultrasonication was carried out using an ultrasonic bath USC1200TH (Prolabo; inner dimension: 300 mm  240 mm  200 mm) with an electrical power of 400 W (i.e. acoustic power of 1 W/cm2), maximal heating power of 400 W and variable

frequencies, equipped with a digital timer, a frequency and a temperature controller. 2.4. Extraction A 50 mg lyophilized and milled seedcoat was extracted in 10 mL of the selected extraction solvent (methanol, ethanol, butanol or water) and homogenized using a blender (Ultraturrax, T25 basic) for 1 min at 19,000 rpm. The extraction was conducted in the ultrasonic bath described in the Section 2.3. during an extraction time ranging from 0 to 60 min at an operating temperature ranging from 25 °C to 60 °C. The position in the ultrasonic bath was chosen according to the Aluminum Foil Efficiency Test. Following extraction, the extract was centrifuged for 15 min at 3,000 rpm and the supernatant was filtered through 0.45 lm nylon syringe membranes prior to HPLC injection. To test the efficiency of the USAE method developed herein, this method was compared with the conventional heat reflux method described by Eliasson et al. [7], the microwave assisted extraction method described by Beejmohun et al. [2] and the enzyme assisted method, using cellulase R10 from Trichoderma reesei described by Renouard et al. [13]. 2.5. HPLC The quantification of compounds was carried out on a Varian liquid chromatographic system as described in [4]. The separation was performed at 35 °C on a Purospher (Merck) RP-18 column (250  4.0 mm i.d.; 5 lm). Detection was performed at 280 nm. Compounds were identified by comparison of their retention times and UV spectra to those of authentic standards. Quantification was performed using calibration curves of each standards ranging from 0.0125 to 0.5 mg/ml with a correlation coefficient of at least 0.999 (Table 1) and using o-coumaric acid as an internal standard (at a 0.05 mg/mL final concentration in the extract). The compounds were quantified using a Varian liquid chromatographic system composed of Varian Prostar 230 pump, Metachem Degasit, Varian Prostar 410 autosampler and Varian Prostar 335 Photodiode Array Detector (PAD) and driven by Galaxie version 1.9.3.2 software. The separation was performed at 35 °C on a Purospher (Merck) RP-18 column (250  4.0 mm i.d.; 5 lm). The mobile phase was composed of 0.2% acetic acid in water (solvent A) and methanol (solvent B). A nonlinear gradient was applied for the mobile phase variation with a flow rate of 0.8 ml/min as follows: from 0 to 40 min of A–B: 90:10 (v/v) to 30:70 (v/v), from 41 to 50 min of A–B: 30:70 (v/v) to 0:100 (v/v), and A–B: 0:100 (v/v) from 51 to 60 min. 2.6. LC–ESI–MS LC–MS analyses were performed on a Waters 2695 Alliance coupled with a single quadrupole mass spectrometer ZQ (Waters-Micromass, Manchester, UK) equipped with an electrospray ion source (ESI–MS) as described in [4]. Data were collected in the centroid mode. Data acquisition and processing were performed with MassLynx V4.0 software. Same conditions as described above were employed for the chromatography. A sample

Table 1 Validation parameters of the HPLC method.

SDG HDG FerG CafG p-CouG

Retention time (mn)

Regression equation

R2

LOD (ng)

LOQ (ng)

Precision (% RSD)

Stability (% RSD)

Repeatability (% RSD)

Recovery (%)

26.98 25.05 24.67 17.08 19.40

y = 0.426 y = 1.430 y = 0.354 y = 0.902 y = 0.185

1.0000 0.9994 0.9991 0.9993 0.9996

2.5 4.3 5.4 5.2 8.8

8.0 14.0 17.3 16.6 28.2

0.66 0.42 0.56 0.54 1.23

0.82 0.26 0.78 0.65 1.44

0.45 0.98 0.67 0.63 0.78

101.3 99.7 97.8 95.3 104.5

x – 1.855 x – 1.31 x – 0.530 x – 1.336 x + 0.917

179

C. Corbin et al. / Ultrasonics Sonochemistry 26 (2015) 176–185

b

EtOH

6

pCouG ***

4

2

d

*

***

* **

0

MeOH

EtOH

nd

ButOH

4

CafG

3

***

3

1

H2O H2O

***

5

nd

0

ButOH

***

P-CouG Content (mg/g DW)

MeOH

2

nd

nd

H2O H 2O

4

*

0

c

***

SDG Content (mg/g DW)

***

5

HDG

*

*

10

6

* ***

***

***

15

8

*

* ***

25 20

SDG

HDG Content (mg/g DW)

30

CafG Content (mg/g DW)

a

2

1

0 H2O H2O

MeOH

ButOH

H2O H2O

MeOH

EtOH

ButOH

3

FerG *** 15 min

2

30 min

**

45 min

1

60 min

***

FerG Content (mg/g DW)

e

EtOH

0 H2O H2O

MeOH

EtOH

ButOH

Fig. 2. Extraction efficiency of water, ethanol, methanol and n-butanol. For each solvent (H2O for water, MeOH for methanol, EtOH for ethanol and ButOH for n-butanol), 4 extraction times have been tested (15, 30, 45 and 60 min.) and results are presented for each compound in mg/g of DW; (a) SDG; (b) HDG; (c) p-CouG; (d) CafG; (e) FerG. *, **, *** indicates statistical differences between two consecutive extraction time for each compound extracted with each solvent with a p value < 0.05, 0.01 and 0.001 respectively.

injection volume of 20 lL (methanol solutions at 0.1 g/L for the reference compound and at 1 g/L for the crude extracts) on the KROMASIL column was carried out for reference compounds and raw cell extracts. The effluent was flow-splitted via a PEEK tee with 1/5 of the flow directed towards the ESI source of the ZQ instrument and the residual 4/5 directed towards a PDA detector (Waters 2996). LC–ESI–MS data were recorded in the positive and negative ion modes. The source and desolvation temperatures were maintained at 120 and 250 °C, respectively. Nitrogen was used as a drying and nebulizing gas at flow rates of 450 and 100 L/h, respectively. The capillary voltage was ±3.5 kV and a cone voltage ranging from ±20 to ±60 V was applied (±ESI). Scanning was performed in the range of 50–1950 Da at a scan rate of 1 s per scan. Data were collected in the centroid mode. Data acquisition and processing were performed with MassLynx V4.0 software. Table 2 Physical characteristics of the solvents used in the present study (provided by suppliers).

Methanol Ethanol n-Butanol Water

Vapor pressure (mmHg)

Surface tension (mN cm 1)

Viscosity (cP)

127.0 59.0 4.2 23.8

22.6 23.7 24.2 72.8

0.55 1.20 2.95 1.00

2.7. Method validation Standard calibration curves were generated using five standard dilutions ranging from 50 to 1000 lg/mL. Arithmetic means of each triplicate were calculated. The linear regression equations were carried out by plotting the peak areas against the injected amounts of standard compounds. The linearity was demonstrated by coefficient of determination (R2). The limits of detection (LOD) and the limits of quantification (LOQ) were determined based on the signal-to-noise ratio (S:N) of approximately 3:1 and 10:1, respectively. Accuracy was evaluated by measuring recovery rates. Flaxseed hulls were homogenized and separated into two parts of equal mass, one of which was spiked with a known volume of stock solutions. The spiked and non-spiked parts were analyzed by HPLC in triplicate following the procedures described. The recovery rates were calculated according to the following formula: Recovery rate = (amount in spiked part amount in non-spiked part)/(spiked amount)  100. The method precision and stability were evaluated by determining the intraday and interday variations respectively, which were calculated from data obtained by the repeated injections of standard solutions. The intraday variation was obtained by five replicates in a day and the interday variation was determined by three injections over three continuous days. Retention times and peak areas were assessed. The precision was further checked by measuring the repeatability using five

180

C. Corbin et al. / Ultrasonics Sonochemistry 26 (2015) 176–185

continuous injections of the same extracted sample. The precision was expressed as the relative standard deviation (RSD, %).

acid Schiff to visualize the cell walls in each tissue before permanent mounting and microscopic observation. Pictures were taken with a CANON EOS 600D digital camera.

2.8. Scanning Electron Microscopy 2.10. Reducing sugar quantification Surface observations of the whole seeds structure by SEM were carried out using a LEO 1430 VP electronic microscope at room temperature in variable pressure mode, allowing the observation of non-conductive samples such as those presented herein with a resolution of several micrometers. In this way, the charged particles drifted away during measurement were unaffected. 2.9. Histology Mucilage release by solvent was appreciated thanks to a 0.1% (w/v) toluidine blue coloration of the solvent. Whole seeds following ultrasonic treatment were stained by shaking in 0.01% (w/v) Ruthenium Red (Sigma–Aldrich) for 2 h and then rinsed three times with distilled water. Samples were fixed in formalin acetic alcohol (10% formalin (37% formaldehyde stabilized with methanol), 5% glacial acetic acid, 60% ethanol) for 24 h, dehydrated and embedded in paraffin. Semi-thin sections (5 lm) were made on a RM 2145 rotary microtome (Leica, Wetzlar, Germany), then deparaffinized and counter-stained with naphthol blue black, safranin O and periodic

Reducing sugars extracted from the seeds were quantified before and following ultrasound treatment of water imbibed seeds, using 3–5 dinitrosalicylicacid (DNS) [16]. DNS acts on reducing sugar in order to produce 3-amino-5-nitrosalicylic acid which concentration can be measured by spectrophotometry at 490 nm. Results were expressed in mg of glucose equivalent per g of seed using a calibration curve of glucose ranging from 0.25 to 2 g/L with a correlation coefficient of 0.9999. 2.11. Statistical treatment of data The means and the standard deviation were used to present the data composed of at least three independent replicates. Student’s t-test was performed for comparative statistical analyses. Significant thresholds at p < 0.05, 0.01 and 0.001 were selected for all statistical tests and represented by *, ** and *** respectively. Model and response surface plots resulting from the combination of extraction duration with ultrasound frequency were conducted using the R package.

a Ethanol

Water

2 mm

2 mm

b

n-Butanol

Methanol

2 mm

Control Seed

2 mm

US-treated seed

mucilage

1 mm

1 mm

c 800

Total reducing sugars (mg/g glc equivalent)

***

600

400

200

0 Water

EtoH Untreated

MeOH

BuOH

US-treated

Fig. 3. Mucilage released and hydrolysis from flax seeds before and after an ultrasonic treatment. (a) Toluidine blue was added in the solvents to evidence the release of the mucilage. (b) Seeds incubated in water were submitted to ultrasonic treatment (1 h; 45 kHz, 25 °C) whereas control seeds underwent only water imbibition. After incubation, seeds were stained by ruthenium red. Mucilage is characterized by the pink pigmentation. (c) Total reducing sugar released from mucilage following US-treatment in the different solvents.

181

C. Corbin et al. / Ultrasonics Sonochemistry 26 (2015) 176–185

3. Results and discussion 3.1. Evaluation of the extraction capacities of different solvents for UAE of flaxseed phenolics Probes and bath constitute the two most common systems for ultrasonic application. Ultrasonic probe immersed directly in the solution and therefore constantly in contact with the sample makes reproducibility and repeatability difficult, and increases the risk of sample contamination and degradation [17]. On the

a

contrary, bath sonicators act on a wide range of samples simultaneously with a higher reproducibility [17]. Therefore, we chose to carry out UAE using a bath sonicator. Since acidic hydrolysis of flax seed extract is a destructive and less productive method [9], the alkaline hydrolysis was employed. We have previously demonstrated that alkaline hydrolysis with 0.1 N sodium hydroxide was sufficient to release SDG from the macromolecular complex [2,18]. Typical HPLC chromatogram of a flaxseed sample extracted by ultrasound treatment in water supplemented with 0.1 N sodium hydroxide is shown in Fig. 1c. The

Mucilaginous cells

Sclerous cells

OUTER SEEDCOAT

Parenchymatous cells

Compressed cells

Endosperm cells 20μm

b

c

INNER SEEDCOAT

Brown cells

Zone 1

Zone 1 300μm

20μm

e

d

Zone 1’

Zone 2’ Zone 1’ Zone 3’ 300μm

f

20μm Zone 2’

20μm

g

Zone 3’

20μm

Fig. 4. Effect of ultrasound on the flaxseed ultrastructure. (a) Anatomy of a mature flaxseed coat. Micrograph of a semi-thin section (5 lm) of mature flaxseed (cv. Barbara) coats stained with naphthol blue black, safranin O and periodic acid Schiff. Bar represents 50 lm. (b) Seeds were incubated in water and submitted to ultrasound treatment (1 h; 45 kHz, 25 °C) whereas control seeds underwent only water imbibition. Representative Scanning Electron Microscopy (SEM) surface observations of whole control seeds are presented in (b) and the magnification of the zone 1 in (c) following by whole ultrasound-treated seeds in (d) and 3 magnifications of different zones in (e–g). Bar is 300 lm for whole seeds and 20 lm for magnifications.

C. Corbin et al. / Ultrasonics Sonochemistry 26 (2015) 176–185

Immunolocalization data have suggested that the accumulation of SDG complex is mainly restricted to the outer integument of mature flaxseed [23]. Our present results demonstrated that mucilage degradation through US-treatment could be interesting for the lignan and other phenolic compounds extraction from the complex. These observations are consistent with the already demonstrated capacity of ultrasounds to extract mucilage from Salvia macrosiphon seeds at a 20 kHz frequency [24]. Here, considering the parameters employed in this study, a complete degradation of mucilaginous cell layer probably occurred thus reducing the possibility of phenolics entrapment by this carbohydrate polymer. 3.2. Development and validation of the UAE of flaxseed phenolics The influence of NaOH concentration (Fig. 5a) and extraction temperature (Fig. 5b) was then evaluated using water as extraction solvent. In order to analyze the relative influence of these two parameters, a non-optimal duration of 30 min and a fixed US frequency of 30 kHz were chosen. Best extraction yield was obtained with 0.2 N NaOH and an incubation temperature of 25 °C. In previous works, NaOH concentration has been evidenced as a determinant parameter for the extraction yield of hydroxycinnamic acid glucosides [2,7], SDG [2] and HDG [4] from flaxseed. The decrease in extraction yield observed at higher temperature could result from a degradation of the compound at high temperature as previously observed for HDG using microwave-assisted extraction (MAE) [4]. Extraction time appeared to be an important parameter because an improvement of the yield was observed when extending the extraction time. According to the literature, another important parameter when performing UAE is the frequency used for the extraction. Therefore, the extraction yield for each compound

20 18 16 14 12 10 8 6 4 2 0

SDG HDG FerG pCouG

*

Quanty (mg/g DW)

a

*

0,1N

CafG

0,2N

0,5N

NaOH Concentraon

b

20 18 16 14

* SDG

10

HDG

***

12 8

FerG

**

6

*

4

pCouG

**

*

** **

glucosides of caffeic acid (m/z = 341.4 [M H] ), p-coumaric acid (m/z = 327.3 [M H] ) and ferulic acid (m/z = 354.3 [M H] ) appeared at 17.08, 19.40 and 24.67 min, respectively, whereas the flavonol HDG (m/z = 625.1 [M H] ) and lignan SDG (m/z = 685.8 [M H] ) eluted at 25.05 and 26.98 min respectively (Fig. 1c). Flaxseed phenolics are usually extracted using methanol as solvent, pure or in mixture with other solvents as recently described by Teh and Birch [19]. However, due to its acute toxicity, other solvents would be preferred for most applications such as food or cosmetic industries [20]. Here, the ultrasound-assisted extraction (UAE) efficiency kinetic of methanol was first compared to the extraction efficiency of other less-toxic solvents such as ethanol, n-butanol and water at a working frequency of 45 kHz. Extraction time for each solvent appeared as important parameters for UAE of flaxseeds phenolics (Fig. 2). Water displayed the best efficiency for each phenolic compound followed by methanol, while ethanol and n-butanol appeared to have poorer extraction capacities (Fig. 2). It is well accepted that the physical characteristics of the solvent greatly affect cavitation and bubble collapse. Physical characteristics (provided by the suppliers) of the solvents used in the present study are listed in the Table 2. Cavities are more readily formed when using a solvent with high vapor pressure, low viscosity and low surface tension but it is also accepted that high vapor pressure results in more vapors entering the cavitation bubble during its formation and its collapse is cushioned and less energetic [21]. Another important parameter to take into account is the solubility of the extracted compound in the solvent chosen. The glycosides studied here represent the water soluble storage forms of these phenolics [18]. The extraction of a compound from plant material is directly related to its polarity matching with the extraction solvent [22]. All these considerations could explain that the best extraction results were obtained with water as solvent. Eliasson et al. [7] reported differences in the yield of SDG obtained by different groups working with the same flaxseed cultivar (Barbara), one hypothesis being that these differences are due to inefficient extraction of oligomers from flaxseed matrix. Indeed, mucilage has been shown to play a critical role in the lignan trapping phenomenon during the extraction process [13] and mucilage hydrolysis using hydrolases resulted in an enhanced release of (+)-secoisolariciresinol from flaxseed hulls [13]. Whole seeds incubation in the four tested solvents (water, ethanol, methanol and n-butanol) for 1 hour prior to a solvent coloration using toluidine blue revealed their respective capacity to release the mucilage (Fig. 3a). As shown in Fig. 3a, water displayed the best capacity for mucilage release whereas other tested solvents did not allow mucilage observation. These observations were confirmed by the ruthenium red staining giving a pink pigmentation to the mucilage (Fig. 3b). In comparison to the control seeds, where the mucilage was clearly visible, ultrasound-treated seeds exhibited no apparent mucilage degradation from the seed coat (Fig. 3b) indicating a possible hydrolysis of this polymer. The quantification of the total reducing sugar liberated following US-treatment in water also supported this hypothesis (Fig. 3c). Flax seed integument is organized from the outside to the inside, in an outer integument (with the mucilaginous cell, parenchymatous cell, sclerified cell and compressed cell layers), and an inner integument (comprising the brown cell and endosperm cell layers) (Fig. 4a). Seeds examination under Scanning Electron Microscopy (SEM) evidenced the destruction of the first cell layers of the seed coat in ultrasound-treated cells (Fig. 4b–g). The magnifications of the 3 zones of the ultrasound-treated seed showed damages of the cells (Fig. 4e), cracks in the cell layers of the seed coat (Fig. 4f) and cells wrenching from superficial cell layers (Fig. 4g) compared to the untreated seeds where the integrity of the external cell layers can undoubtedly be noticed (Fig. 4c).

Quanty (mg/g DW)

182

2

CafG

0 25°C

40°C

60°C

Temperature Fig. 5. Effect of sodium hydroxide concentration and temperature on UAE of phenolics. Quantity in mg/g DW extracted for each compound are presented in (a) at room temperature with 3 different NaOH concentrations (0.1, 0.2 or 0.5 N) and in (b) with a 0.5 N NaOH at 3 different temperatures (25 °C, 40 °C or 60 °C). Statistical differences were determined for each compound between two consecutive NaOH concentrations and results are indicated by *, **, *** corresponding to a p value < 0.05, 0.01 and 0.001 respectively.

183

C. Corbin et al. / Ultrasonics Sonochemistry 26 (2015) 176–185

resulting from a combination of the extraction time (15, 30 and 60 min) with the ultrasound frequency (15, 30 and 45 kHz) parameters was determined and the model describing these effects is presented as response surface plot in Fig. 6. The extraction time and the ultrasound power displayed a positive effect on the extraction yield as the response surface representations reached a plateau for the maximum values of these two parameters for the hydroxycinnamic acid glucosides (Fig. 6b, c and d, ESM1, 2). These results confirmed the influence of the extraction time on the extraction yield, as previously described for the UAE of other natural products [25–27] Nevertheless, the combination of duration and power maximum values has a detrimental effect on SDG and HDG yield (Fig. 6a and b). A similar decrease in HDG extraction yield from flaxseed cake with prolonged extraction time at high microwave power has been previously described [4]. It is accepted that sample degradation could occur with non-optimal sonication conditions [17] and decrease in extraction yield with higher ultrasound frequency was already observed in salvianolic acid B extraction from Salvia miltiorrhiza roots [26]. Once again, the observed decrease in extraction yield could indicate a degradation of the compound for prolonged extraction duration at high ultrasound frequency. Considering the above results, the optimal extraction conditions for UAE of lignan and other phenolic compounds from flaxseeds were: water as solvent supplemented with 0.2 N of NaOH for the

alkaline hydrolysis of the SDG–HMG complex, an extraction time of 60 min at a temperature of 25 °C and an ultrasound frequency of 30 kHz. In order to ensure the precision, accuracy, stability and repeatability of the phenolics UAE and the HPLC-PAD method, a validation was undertaken. Representation of the peak area and standard concentrations revealed high linear correlations in the range of 50–1000 lg/mL. The linear regression of the 5-point calibration graph showed a R2-value > 0.999 and the slope of the standards covering the analytical range varied at most 1% relative standard deviation (RSD) over a four weeks period. The Table 1 showed the limits of detection (LOD, S/N = 3), the limits of quantification (LOQ, S/N = 10) and the precision, stability and repeatability results of the optimized extraction protocol. The determination of the instrumental precision was realized by five injections of the same sample. A RSD from 0.42% to 1.23% for the chromatographic method proved its precision. Application of the whole extraction procedure three times to the same batch of material allowed the evaluation of the repeatability where the obtained RSD value was low (0.41–0.98%; Table 1). The same sample was injected six times (0, 6, 12, 24, 48 and 72 h after it was prepared) to determine the stability. The small observed value for the RSD, from 0.26% to 1.44% (Table 1) confirmed the good stability of the extracted sample. The separation method accuracy measurement was assessed with standard addition at one level of 5 lg/mL and a good recovery

a

b 25

SDG

7

HDG

6 20 HDG content (mg/g DW)

SDG content (mg/g DW)

25 20 15

15

10

6

5 4

4 2

5 45

45

10

60

3

60 30

30

40

40 15 20

15

2

20

5 1

5

pCouG

d

CafG

3,5

45

2

2,0 CafG content (mg/g DW)

p-CouG content (mg/g DW)

3 2

3

2,5

2 2,0 1 45

1,5

60 30

40 20

2

1,5

1 45 60

60

30 15

2,5

FerG

3,0

4 4

e FerG content (mg/g DW)

c

40 1 15

1,0

30

1,0

40 15 20

20 0,5

0,5

Fig. 6. Response surface plots of UAE from flaxseed hulls showing the effect of both ultrasound frequency and extraction time on the content of SDG (a), HDG (b), p-CouG (c), CafG (d) and FerG (e).

184

C. Corbin et al. / Ultrasonics Sonochemistry 26 (2015) 176–185

Table 3 Re-extraction data.

SDG HDG FerG CafG pCouG

First extraction (mg/g DW)

Second extraction (mg/g DW)

Third extraction (mg/g DW)

24.07 ± 0.70 6.84 ± 0.33 2.58 ± 0.04 3.63 ± 0.14 5.04 ± 0.23

2.11 ± 0.40⁄⁄⁄ 1.40 ± 0.23⁄⁄⁄ 0.27 ± 0.05⁄⁄⁄ 0.66 ± 0.15⁄⁄⁄ 0.41 ± 0.07⁄⁄⁄

nd⁄⁄⁄ nd⁄⁄⁄ nd⁄⁄⁄ nd⁄⁄⁄ nd⁄⁄⁄

Values are the mean ± SD of 4 independent replicates; nd: not detected; ⁄, corresponding to a p value < 0.05, 0.01 and 0.001 respectively.

⁄⁄ ⁄⁄⁄

,

of the compounds ranging from 95.3 to 107.5 (Table 1) was noticed. Furthermore, the procedure of re-extraction consisted in hulls submission to 3 sequential extraction procedures. The results revealed a phenolic detection with the second extraction meaning that after the first extraction a small un-extracted quantity remained (Table 3). No compound was detected after a third extraction (Table 3).

3.3. Comparison of the UAE with other methods Phenolic compounds from flaxseed are stored esterified in a complex form and their extraction was usually achieved by alcoholic solid–liquid extraction and alkaline treatment [7,8,18]. Acidic hydrolysis was also described in the literature but with the major drawback of being destructive [9,12]. Microwave-assisted extraction recently described allows a gain of time and a slight improvement in terms of yield [2,4,10,28]. Various enzymes were also tested for enzyme-aided extraction such as ß-glucosidase, sulfatase or cellulase [11–13,29,30]. Such extraction methods allow the recovery of unique (aglycone) products with relative high yield but are time consuming. Table 4 presents the comparison of the present UAE method with other protocols such as the conventional heat reflux extraction described by Eliasson et al. [7], the microwave-assisted extraction designed by Beejmohun et al. [2] and Fliniaux et al. [4] and the cellulase-aided extraction proposed by Renouard et al. [13]. The results of the cellulase-aided extraction in the different aglycones were here expressed in glucoside equivalent contents per gram of whole seed DW in order to enable comparison. The results clearly indicated that UAE gave an overall higher yield, particularly for SDG with 23.6 mg/g DW and HDG with 5.9 mg/g DW, and in a shorter extraction time compared with the conventional method or the enzyme-aided extraction (Table 4). Levels of SDG obtained

Table 4 Comparison of different extraction methods for flaxseed phenolics extraction from the macromolecular complex.

with whole flaxseed in this study were in the top of the range of those reported in the literature. Liggins et al. [31] achieved a similar yield for SDG (from 14.8 to 25.5 mg/g DW) but under anhydrosecoisolariciresinol form. The major drawback of this compound is its conversion into a new mammalian lignan (enterofuran) by the microbiota which health benefit has not already been demonstrated [32]. Considering SDG, FerG and CafG, Eliasson et al. [7] work displayed the same range of extraction yield by using different flaxseed cultivars as starting materials. Same observation can be noticed for HDG [4]. Depending on the flaxseed cultivar, the SDG content varied considerably [3,7], so only results obtained with the same cultivar can be compared. The Barbara cultivar was the most studied cultivar, mainly for its high SDG accumulation. The different yields for SDG extraction (extrapolated in SDG content in whole seed DW for comparison) from this cultivar were reported in the literature: 11.8 mg/g by Charlet et al. [33] based on acidic hydrolysis, 11.4 mg/g by Johnson et al. [8] using a sequential alkaline extraction, 17.3 mg/g by Eliasson et al. [7] after a direct alkaline extraction, 9.8 mg/g by Beejmohun et al. [2] using a microwave-assisted extraction and 20.5 mg/g by Renouard et al. [13] thanks to a cellulase-aided extraction. Our SDG content of 23.6 mg/g extracted with ultrasound from the cultivar Barbara is thus higher than those described in other published methods (Table 4).

4. Conclusion Ultrasound-assisted extraction (UAE) lignan and other phenolic compounds of flaxseed was investigated for the first time. This method was found to be very efficient for the reduction of mucilage entrapment of these phenolics thus allowing a high extraction yield. Parameters affecting UAE were optimized and considering these results, the optimal extraction conditions for UAE of lignan and other phenolic compounds from flaxseeds were: water as solvent supplemented with 0.2 N of sodium hydroxide for the alkaline hydrolysis of the SDG–HMG complex, an extraction time of 60 min at a temperature of 25 °C and an ultrasound frequency of 30 kHz. Under these optimized and validated conditions, highest yields of SDG, HDG and hydroxycinnamic acids were detected in regards to other published methods. The process presented herein is a valuable method for efficient extraction and quantification of the main flaxseed phenolics. Moreover this UAE is of particular interest within the context of green chemistry as it uses water as solvent, without the production of contaminants, allows reducing energy consumption by using innovative technology and valorization of by-products since flaxseed cakes result from the production of flax oil.

Acknowledgements This project was funded by the Natur’ACTIV project (HLB1307) from the Région Centre, the Conseil Général d’Eure et Loir and the Ligue Contre le Cancer, Comité d’Eure et Loir. CC obtained a PhD grant from the Région Centre. EAL obtained a post-doctoral grant from the Région Centre (Natur’ACTIV project). We wish to thank Jean Paul Trouvé and the Coopérative Linière Terre de Lin for the donation of flaxseed used in this study.

Values are the mean ± SD of 4 independent replicates. UAE: ultrasound-assisted extraction, MAE: microwave-assisted extraction, EAE: enzymatic-assisted extraction, heat reflux: conventional alkaline extraction described by Eliasson et al. [7]; the same letter indicates that mean values in the different organs for the considered species are not statistically different (P > 0.05).

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ultsonch.2015.02. 008.

C. Corbin et al. / Ultrasonics Sonochemistry 26 (2015) 176–185

References [1] B.D. Oomah, Flaxseed as a functional food source, J. Sci. Food Agric. 81 (2001) 889–894. [2] V. Beejmohun, O. Fliniaux, E. Grand, F. Lamblin, L. Bensaddek, P. Christen, et al., Microwave-assisted extraction of the main phenolic compounds in flaxseed, Phytochem. Anal. 18 (2007) 275–282. [3] N. Westcott, A. Muir, Flax seed lignan in disease prevention and health promotion, Phytochem. Rev. 2 (2003) 401–417. [4] O. Fliniaux, C. Corbin, A. Ramsay, S. Renouard, V. Beejmohun, J. Doussot, et al., Microwave-assisted extraction of herbacetin diglucoside from flax (Linum usitatissimum L.) seed cakes and its quantification using an RP-HPLC–UV system, Molecules 19 (2014) 3025–3037. [5] A. Kosinska, K. Penkacik, W. Wiczkowski, R. Amarowicz, Presence of caffeic acid in flaxseed lignan macromolecule, Plant Foods Hum. Nutr. 66 (2011) 270– 274. [6] K. Schoenrock, U., Untiedt, S., Kux, U. and Inoue, Application of ferulic acid glucosides as anti-irritants in cosmetic and topical dermatological preparations, 1997. [7] C. Eliasson, A. Kamal-Eldin, R. Andersson, P. Aman, High-performance liquid chromatographic analysis of secoisolariciresinol diglucoside and hydroxycinnamic acid glucosides in flaxseed by alkaline extraction, J. Chromatogr. A 1012 (2003) 151–159. [8] P. Johnsson, a. Kamal-Eldin, L.N. Lundgren, P. Aman, HPLC method for analysis of secoisolariciresinol diglucoside in flaxseeds, J. Agric. Food Chem. 48 (2000) 5216–5219. [9] H.-B. Li, C.-C. Wong, K.-W. Cheng, F. Chen, Antioxidant properties in vitro and total phenolic contents in methanol extracts from medicinal plants, LWT – Food Sci. Technol. 41 (2008) 385–390. [10] W. Zhang, S. Xu, Microwave-assisted extraction of secoisolariciresinol diglucoside from flaxseed hull, J. Sci. Food Agric. 87 (2007) 1455–1462. [11] I.E.J. Milder, I.C.W. Arts, D.P. Venema, J.J.P. Lasaroms, K. Wähälä, P.C.H. Hollman, Optimization of a liquid chromatography tandem mass spectrometry method for quantification of the plant lignans secoisolariciresinol, matairesinol, lariciresinol, and pinoresinol in foods, J. Agric. Food Chem. 52 (2004) 4643–4651. [12] T. Kraushofer, G. Sontag, Determination of some phenolic compounds in flax seed and nettle roots by HPLC with coulometric electrode array detection, Eur. Food Res. Technol. 215 (2002) 529–533. [13] S. Renouard, C. Hano, C. Corbin, O. Fliniaux, T. Lopez, J. Montguillon, et al., Cellulase-assisted release of secoisolariciresinol from extracts of flax (Linum usitatissimum) hulls and whole seeds, Food Chem. 122 (2010) 679–687. [14] M.A. Rostagno, M. Palma, C.G. Barroso, Ultrasound-assisted extraction of soy isoflavones, J. Chromatogr. A 1012 (2003) 119–128. [15] J. Wu, L. Lin, F.T. Chau, Ultrasound-assisted extraction of ginseng saponins from ginseng roots and cultured ginseng cells, Ultrason. Sonochem. 8 (2001) 347–352. [16] G.L. Miller, Use of dinitrosaiicyiic acid reagent for determination of reducing sugar, Anal. Chem. 31 (1959) 426–428. [17] A. Khoddami, M.A. Wilkes, T.H. Roberts, Techniques for analysis of plant phenolic compounds, Molecules 18 (2013) 2328–2375.

185

[18] C. Hano, I. Martin, O. Fliniaux, B. Legrand, L. Gutierrez, R.R.J. Arroo, et al., Pinoresinol-lariciresinol reductase gene expression and secoisolariciresinol diglucoside accumulation in developing flax (Linum usitatissimum) seeds, Planta 224 (2006) 1291–1301. [19] S.S. Teh, E.J. Birch, Effect of ultrasonic treatment on the polyphenol content and antioxidant capacity of extract from defatted hemp, flax and canola seed cakes, Ultrason. Sonochem. 21 (2014) 346–353. [20] F. Kerton, R. Marriott, Alternative Solvents for Green Chemistry, Royal Soc., 2013. [21] S. Vajnhandl, A. Majcen Le Marechal, Ultrasound in textile dyeing and the decolouration/mineralization of textile dyes, Dyes Pigm. 65 (2005) 89–101. [22] J. Wang, B. Sun, Y. Cao, Y. Tian, X. Li, Optimisation of ultrasound-assisted extraction of phenolic compounds from wheat bran, Food Chem. 106 (2008) 804–810. [23] J. Attoumbré, C. Bienaimé, F. Dubois, M.-A. Fliniaux, B. Chabbert, S. Baltora-Rosset, Development of antibodies against secoisolariciresinol–appli cation to the immunolocalization of lignans in Linum usitatissimum seeds, Phytochemistry 71 (2010) 1979–1987. [24] A. Farahnaky, Sh. Bakhshizadeh-Shirazi, Gh. Mesbahi, M. Majzoobi, E. Rezvani, G. Schleining, Ultrasound-assisted isolation of mucilaginous hydrocolloids from Salvia macrosiphon seeds and studying their functional properties, Innov. Food Sci. Emerg. Technol. 20 (2013) 182–190. [25] E.-Q. Xia, X.-X. Ai, S.-Y. Zang, T.-T. Guan, X.-R. Xu, H.-B. Li, Ultrasound-assisted extraction of phillyrin from Forsythia suspensa, Ultrason. Sonochem. 18 (2011) 549–552. [26] J. Dong, Y. Liu, Z. Liang, W. Wang, Investigation on ultrasound-assisted extraction of salvianolic acid B from Salvia miltiorrhiza root, Ultrason. Sonochem. 17 (2010) 61–65. [27] E.-Q. Xia, Y.-Y. Yu, X.-R. Xu, G.-F. Deng, Y.-J. Guo, H.-B. Li, Ultrasound-assisted extraction of oleanolic acid and ursolic acid from Ligustrum lucidum Ait, Ultrason. Sonochem. 19 (2012) 772–776. [28] S. Nemes, V. Orsat, Microwave-assisted extraction of secoisolariciresinol diglucoside-method development, Food Bioprocess Technol. 4 (2011) 1219– 1227. [29] W. Mazur, T. Fotsis, K. Wähälä, S. Ojala, A. Salakka, H. Adlercreutz, Isotope dilution gas chromatographic-mass spectrometric method for the determination of isoflavonoids, coumestrol, and lignans in food samples, Anal. Biochem. 233 (1996) 169–180. [30] W.R. Obermeyer, S.M. Musser, J.M. Betz, R.E. Casey, A.E. Pohland, S.W. Page, Chemical studies of phytoestrogens and related compounds in dietary supplements: flax and chaparral, Proc. Soc. Exp. Biol. Med. 208 (1995) 6–12. [31] J. Liggins, R. Grimwood, S.A. Bingham, Extraction and quantification of lignan phytoestrogens in food and human samples, Anal. Biochem. 287 (2000) 102– 109. [32] K. Struijs, J.-P. Vincken, H. Gruppen, Bacterial conversion of secoisolariciresinol and anhydrosecoisolariciresinol, J. Appl. Microbiol. 107 (2009) 308–317. [33] S. Charlet, L. Bensaddek, S. Raynaud, F. Gillet, F. Mesnard, M.-A. Fliniaux, An HPLC procedure for the quantification of anhydrosecoisolariciresinol. Application to the evaluation of flax lignan content, Plant Physiol. Biochem. 40 (2002) 225–229.