Ultrasound-Assisted Extraction and Biological Activities of Extracts of Brassica oleracea var . capitata

Brassica sp. is a vegetable belonging to the family Brassicaceae, which is one of the most important families cultivated worldwide. Some species of this family are of pharmacological interest. Several studies have revealed that they exhibit anti-infl ammatory, antimycotic, photoprotective, antihyperglycemic, anticarcinogenic and antioxidant activities (1–3). Among the various Brassica species, Brassica oleracea L. var. capitata (cabbage) has been widely studied because of its biological activities (4). Several authors have reported antioxidant activities of extracts of Brassica against diff erent radicals. Among the methods for testing the ability of some compounds to act as free radical scavengers is the 1,1-diphenyl-2-picrylhydrazyl (DPPH) assay. This procedure has been widely used to evaluate the antiradical activity of several plants, including extracts of Brassica sp., because it is considered to be an easy, precise and reproductive test for evaluating the antioxidant activity (5–7). The antioxidant potential of Brassica extracts against superoxide radicals (O2 ̇), which can cause indirect damage because they lead to the forISSN 1330-9862 scientifi c note


Introduction
Brassica sp. is a vegetable belonging to the family Brassicaceae, which is one of the most important families cultivated worldwide.Some species of this family are of pharmacological interest.Several studies have revealed that they exhibit anti-infl ammatory, antimycotic, photoprotective, antihyperglycemic, anticarcinogenic and antioxidant activities (1-3).Among the various Brassica species, Brassica oleracea L. var.capitata (cabbage) has been widely studied because of its biological activities (4).
Several authors have reported antioxidant activities of extracts of Brassica against diff erent radicals.Among the methods for testing the ability of some compounds to act as free radical scavengers is the 1,1-diphenyl-2-picrylhydrazyl (DPPH) assay.This procedure has been widely used to evaluate the antiradical activity of several plants, including extracts of Brassica sp., because it is considered to be an easy, precise and reproductive test for evaluating the antioxidant activity (5-7).The antioxidant potential of Brassica extracts against superoxide radicals (O 2 ˙-), which can cause indirect damage because they lead to the for-mation of hydroperoxides through the autoxidation of unsaturated fa y acids, was also investigated (5,8,9).However, the use of Brassica extracts against the peroxyl radical (ROO˙), which is an important intermediary in lipid peroxidation, has not been considered much.Specifically, Zhou and Yu (10) evaluated the scavenging activities of vegetables, including broccoli and kale, against DPPH, ROO˙ and O 2 ˙, whereas Ou et al. (11) determined the antioxidant activity of white cabbage and broccoli extracts against peroxyl radicals.
However, the biological activity of extracts obtained from plant species is related to the extraction method because each method aff ects the selectivity of the compounds.In this way, novel extraction techniques, such as the ultrasound-assisted method, have been used to obtain plant extracts.The use of ultrasound can increase cell wall destruction, cause leakage of cellular material, enhance the penetration of solvent into plant cells, facilitate hydration and swelling, and improve mass transfer.These phenomena can increase the extraction of antioxidants while signifi cantly reducing the extraction time, thus improving overall effi ciency.Ultrasound-assisted extraction (UAE) is versatile and can be performed on small and large scale (14-16 Therefore, the objective of this study is to evaluate the antioxidant and antimicrobial activities of Brassica oleracea var.capitata (white cabbage) extracts obtained using ultrasound-assisted extraction.First, the extraction conditions were optimized, and then the extracts were submitted to diff erent hydrolysis conditions before their use in the biological evaluations.The crude and hydrolysed extracts were characterized using gas chromatography coupled with mass spectrometry (GC-MS).

Ultrasound-assisted extraction
Experiments were performed in a reactor with a thermostatic water bath (temperature accuracy of ±1.0 °C).The experimental setup consisted of an ultrasonic bath (model USC 1800A, Unique Inc., Indaiatuba, SP, Brazil) equipped with a transducer with longitudinal vibrations.The ultrasonic unit had an operating frequency of 40 kHz and a maximum ultrasound power output of 132 W. The ultrasonic transducer (surface area of 282.2 cm 2 ) was fi ed at the bo om of the bath horizontally along the length of the bath, corresponding to an ultrasonic intensity of 0.46 W/cm 2 .
For the optimization of the extraction conditions, a central composite rotational design (CCRD) with two independent variables was constructed to evaluate the infl uence of ethanol fraction and temperature on the extraction (21).Table 1 presents the investigated range for each independent variable.The extractions were performed in the ultrasonic bath for 120 min using an Erlenmeyer fl ask containing 2 g of dry material and 10 mL of an aqueous solution of ethanol.The same experiment was performed in duplicate in the presence and absence of ultrasound irra diation for comparison.In this paper, the evaluated response was the sum of the area of all compounds identifi ed by HPLC.For comparison, an extraction was performed by substituting ethanol with methanol under the optimized experimental conditions because there is a difference in the polarity of these two solvents.

Hydrolysis of the crude extracts
The hydrolysis of the crude extracts (methanolic and ethanolic) obtained using the best experimental conditions of the previous step was performed following the method proposed by Robbins et al. (22) with some modifications.Acid hydrolysis consisted of treating 2 mL of the crude extract with 1 mL of HCl (2 M) and then heating the mixture to 95 °C for 1.5 h.Alkaline hydrolysis consisted of treating 5 mL of the extract with 5 mL of an alkaline solution (2 M NaOH, 10 mM EDTA and 1 % ascorbic acid).The reaction mixture was maintained at 30 °C under stirring for 30 min.Next, the pH was adjusted to 3.0 with 8 M HCl solution, and extraction with ethyl acetate (two 5-mL portions) using sonication for 20 min was performed.Organic solvent was removed using a rotary evaporator.The solid residue was resuspended in 2 mL of methanol and fi ltered (polyvinylidene fl uoride, particle size 0.22 µm).Prior to testing for antioxidant and antibacterial activities, these extracts were neutralized.

Chromatographic analysis
In the extraction step, high-performance liquid chromatography (HPLC) was employed to separate the compounds according to the method proposed by Leoni et al. (23) with some modifi cations.For this purpose, a Dionex HPLC (model P680, UV-VIS detector UVD-170 Dionex GmbH, Germering, Germany) equipped with a column ODS-18 (250 mm×5 µm, Thermo Scientifi c, Waltham, MA, USA) using a CH 3 CN/H 2 O 15:85 mixture as the eluent was used.A fl ow rate of 1.0 mL/min, wavelengths of 240 and 254 nm and a total run time of 15 min were employed.
The identifi cation of compounds present in the crude and hydrolysed extracts was performed only in the optimized run of the experimental design (methanolic and ethanolic extracts) using gas chromatography coupled with mass spectrometry detector (GC-MS).The extracts were analyzed with a gas chromatograph (HP 6890, Agilent, Ramsey, MN, USA) interfaced with a mass selective detector (HP 5973, Agilent) with an automatic injection system (HP 6890, Agilent), using a capillary column HP--5ms (30 m×0.32 mm×0.25 µm).Helium was the carrier gas with a fl ow rate of 2 mL/min at a pressure of 0.35 bar and electronic impact mode of 70 eV.Samples of 1 µL were injected at 250 °C interface temperature with the fol-lowing column temperature gradient programming: 70 °C for 1 min, then 12 °C/min up to 280 °C.

Radical DPPH scavenging activity
The analytical method by Zhao et al. (24), with modifi cations, used to measure the radical scavenging activity of crude and hydrolysed extracts of Brassica oleracea var.capitata against DPPH radical was based on the addition of 1500 µL of extract to 1480 µL of DPPH solution plus 20 µL of hydroethanolic solution.A blank assay was performed using 1500 µL of hydroethanolic solution instead of the extract.The resulting solution was le to rest for 30 min.Then, the absorbance of samples was determined at 522 nm using UV-VIS 8453 Hewle -Packard spectrophotometer (Agilent Technologies, Santa Clara, CA, USA).The antiradical activity against DPPH (AA DPPH ) was calculated according to the following equation: /1/ where A DPPH , A and A B are the absorbance of the DPPH solution, sample and blank, respectively.

Superoxide anion radical scavenging activity
The antiradical activity of Brassica oleracea var.capitata extracts, obtained by ultrasound-assisted extraction, against O 2 ˙-radicals was evaluated by the enzymatic system HPX/XOD (24).For this purpose, 100 µL of EDTA (30 mmol/L), 100 µL of HPX (3 mmol/L) and 200 µL of NBT (1.42 mmol/L) were mixed with 100 µL of extract.A er 3 min, 100 µL of enzyme XOD (0.75 U/mL, diluted in phosphate buff er) were added.The fi nal volume of the solution was brought to 3 mL with phosphate buff er (0.05 mol/L, pH=7.4).The blank sample was prepared in the same manner but without the NBT.Additionally, a control test was performed containing all reagents with the solvent employed in the samples and a blank control.After 40 min of reaction, the absorbance of the samples was measured using UV-VIS 8453 Hewle -Packard spectrophotometer (Agilent Technologies) at 560 nm.The antiradical activity against O 2 ˙-(AA O 2 •-) was calculated according to the following equation: /2/ where A and A B are the absorbance of the sample and blank, respectively, and A C and A BC are the absorbance of the control and blank control, respectively.

Peroxyl radical anion scavenging activity
The antiradical activity of Brassica oleracea var.capitata extracts, obtained by ultrasound-assisted extraction, against ROO˙ radicals was evaluated using the fl uorimetric method using DCFH2-DA as a substrate (25).A plate containing 96 wells was subdivided into two regions: the fi rst region corresponded to lines A, B, C and D, and the second region corresponded to lines E, F, G and H.In the fi rst three wells of each region 10 µL of solvent were added to the samples.In the remaining wells of the plate, 10 µL of the extract were added.Then, 127.5 µL of buff er were added to all wells of the plate.Subsequently, 7.5 µL of ultrapure water were added to all wells of region 1, whereas 7.5 µL of ABAP (4 mmol/L) were added to the wells of region 2. Before the analysis, 10 µL of DCFH2-DA (16 µmol/L) were added.A Vitor 2 fl uorimeter (PerkinElmer, Waltham, MA, USA) was programmed to maintain the temperature at 37 °C and to measure the fl uorescence at 485 nm (excitation) and 520 nm (emission) in regular time intervals of 5 min for 30 min.The antiradical activity against ROO˙ (AA ROO˙) was calculated a er 30 min according to the following equation: /3/ where F A and F AB are the fl uorescence of the sample containing ABAP and the fl uorescence of the blank without ABAP, respectively, and F S and F SB are the fl uorescence of the solvent containing ABAP and the fl uorescence of the solvent blank without the ABAP, respectively.

Antibacterial activities (S. aureus and E. coli) of extracts
The antimicrobial activity of crude and hydrolysed extracts of Brassica oleracea var.capitata was determined using the antibiogram with solid-phase (disc-diff usion) methodology.The extracts were resuspended in the extraction solvent (methanol or ethanol) at a concentration of 100 mg/mL.Staphylococcus aureus ATCC 29213 and Escherichia coli ATCC 14948 were inoculated on Müller--Hinton agar and kept in an oven maintained at (36±2) °C for 24 h.As a positive control for S. aureus and E. coli, erythromycin (10 µg/µL) and ciprofl oxacin (0.1 µg/µL), respectively, were applied in the same plate.The antimicrobial activity of the solvents (negative control) was also determined.The degree of sensitivity or resistance of a microorganism was determined by measuring the size of the antimicrobial eff ect zones (size of halo formed).Samples were examined in duplicate against each bacterium.

Statistical analysis
The statistical analysis was performed using ANOVA coupled with the Tukey's test at a 95 % confi dence level using STATISTICA ® v. 8.0.(StatSo Ltd, Tulsa, OK, USA).

Ultrasound-assisted extractions
The process variables studied in this work were temperature in the range of 30-70 °C and ethanol volume fraction in the range of 20-100 %.All experiments were performed in the presence and absence of ultrasound irradiation.Table 1 presents the results obtained with the CCRD concerning the total HPLC area of all compounds separated in the sample, which ranged from 35.44 to 237.90 mAU and from 46.71 to 494.25 mAU in the absence and presence of ultrasound, respectively.Considering the mean value obtained in the 11th run, signifi cant differences (p<0.05) were detected between the amounts obtained in the absence and presence of ultrasound because the mean values were 137.44 and 164.59 units, respectively.
Fig. 1 presents the extraction profi les in the presence and absence of ultrasound expressed as normalized HPLC peak area.The right y-axis shows the normalized HPLC peak area for both experimental sets, black bars show ultrasound extraction and grey without ultrasound, whereas the le y-axis indicates the contribution or increase (white do ed column) in percentage when ultrasound was applied.In all runs, a positive eff ect of ultrasound in comparison with conventional extraction (maceration) is clearly observed, and an average HPLC peak area increase of (36.1±15.5)% was observed.
Note that the ultrasound eff ect can be associated with an increase in the mass transfer process during the reaction through the formation of cavitation bubbles, which provide an important benefi t of opening up the surface of solid substrates.Consequently, at lower temperatures, the energy provided by the ultrasound is responsible for the improvement in the extraction rates, but at higher temperatures thermal energy is suffi cient and the ultrasound exhibits a positive eff ect on the extraction runs.
Based on the results presented in Table 1, the experimental ultrasound conditions for the extraction of bioactive compounds of Brassica oleracea var.capitata were verifi ed in run 5 (30 °C and 60 %, by volume, of ethanol) in the presence of the ultrasound.Considering this result, an additional extraction was performed in which ethanol was substituted with methanol, and the obtained peak area of the extracted sample was 330.5 mAU.In the following steps, only the crude and hydrolysed extracts obtained using methanol or ethanol were considered.

Gas chromatography-mass spectrometry analysis of Brassica oleracea var. capitata extracts
The crude and hydrolysed extracts of Brassica oleracea var.capitata obtained by ultrasound-assisted extraction (at 30 °C and 60 %, by volume, of ethanol or methanol) were analyzed with a gas chromatograph interfaced with a mass selective detector.The composition of the extracts obtained in this work was expressed as the percentage of the normalized peak areas, which are presented in Table 2.It was possible to identify 10 compounds, including derivatives of fa y acids, sulphur compounds and glyco- sides.The crude ethanol was the sample with the highest number of identifi ed compounds.Desulphosinigrin and octadecanoic acid are among the compounds of interest.
Although fl avonoids and glucosynolates, which are usually found in Brassica sp., were not determined in the extracts, a signifi cant amount of glycosides, which are associated with fl avonoids and phenolic acids, were found.The presence of these glycosides can indicate the presence of fl avonoids in the sample.The hydrolysis of the extracts led to diff erences in the chemical profi les of the samples.For example, in both acid-hydrolysed extracts, the presence of 2-furancarboxaldehyde, which is a degradation product formed during the acid hydrolysis of glucose, was detected.The degradation of compounds when ultrasound is employed was well reported in the studies of Pingret et al. (26,27), who stated that acoustic cavitation might be responsible for initiating the formation of degradation products, which can trigger radical chain reactions and provoke substantial quality impairments in those products.The use of ultrasound can intensify the hydrolysis of fl avonoids, leading to the identifi cation of glycosides in the extract.
In addition, in these same samples, the presence of ethyl α--glucopyranoside, which is a compound with a positive eff ect on the prevention of skin barrier disruption (28), was detected.In the crude and alkaline hydrolysis of ethanolic extracts, the presence of desulphosinigrin, a glucosinolate, was observed.Considering the content of fa y acids present in the extracts, hexadecanoic acid was the compound with the highest fraction, 3.35 and 4.77 % in runs 5 and 6, respectively.Fa y acids showed antioxidant (29) and antimicrobial activities (30), which were also evaluated in this study.

Antioxidant activity of ultrasound extracts of Brassica oleracea var. capitata
Table 3 presents the results concerning the antioxidant activities of the crude and hydrolysed extracts of Brassica oleracea var.capitata obtained using ultrasound--assisted extraction (at 30 °C and 60 %, by volume, of eth-anol or methanol) against DPPH, O 2 ˙-and ROO˙ radicals.All samples showed antioxidant activities, which ranged from 13.0 to 80.0, 35.2 to 63.1 and 89.3 to 99.5 % towards DPPH, O 2 ˙-and ROO˙ radicals, respectively.
Concerning the antioxidant activity against DPPH radicals, both the methanolic and ethanolic hydrolysed extracts presented higher activities than the crude extracts.These results were statistically evaluated using the Tukey's test (p<0.05),and the hydrolysed extracts of both methanolic and ethanolic fraction did not present significant diff erences, even though the hydrolysis process was diff erent.
Against the O 2 ˙-radical, the crude methanolic extract had a 1.7-fold higher antioxidant activity than the crude ethanolic extract, which was signifi cantly diff erent (p<0.05).This result indicates that the solvent used in the extraction has an infl uence on the antioxidant activities of the extracts, consistent with the results obtained by Zhao et al. (24), who reported that extracts from a barley variety using 80 % methanol possessed the highest O 2 ˙-scaveng- ing activity, followed by water, 80 % acetone and 80 % ethanol extracts.The hydrolysis of ethanolic extracts had higher activity than the crude extract; however, the acid hydrolysis extracts had a lower activity than alkaline hydrolysis extracts, with the results being statistically signifi cant (p<0.05).Considering the methanolic extracts, the acid hydrolysis extract had a lower activity than the crude extract, but the alkaline hydrolysis extract had a similar activity as the crude extract, where no signifi cant diff erence was detected.
The crude and hydrolysed extracts of Brassica oleracea var.capitata obtained by ultrasound-assisted extraction showed the highest antioxidant activity against ROO˙ radicals, which ranged from 89.3 to 99.5 %.Only the alkaline hydrolysis of methanolic extract showed a signifi cant diff erence compared to other extracts.The high inhibitory activity against ROO˙ radicals verifi ed in all the tested extracts can be a ributed to the fact that this radical is less reactive than the other extracts that have been studied (31).Another possibility is related to the high fractions of the extracts (20 %).
Vrchovská et al. (8) observed the antioxidant activity of aqueous extracts of Brassica oleracea var.costata against DPPH and O 2 ˙-radicals.The antioxidant activity of Brassica oleracea species (white cabbage and broccoli) obtained using diff erent extraction methods, such as acetone/water 50:50 by volume (11), water extraction in combination with hexane extraction (32) and hexane/dichloromethane (1:1 by volume) in combination with acetone/water/acetic acid (70:29.5:0.5 by volume) (33), also presented activities against ROO˙ radicals.Podsedek (34) reported that the contribution of Brassica spp.vegetables to be er health may be related to their antioxidant capacity.
Although several works have reported the antioxidant activities of extracts from diff erent species of Brassica obtained using diff erent extraction methods (5,7), there is no report on common ultrasound method for obtaining hydrolysed extracts.This paper demonstrates that the hydrolysis process can considerably improve the antioxidant activities of the extracts obtained with ultrasound-assisted extraction.For example, the antioxidant activity against DPPH radicals increased from 13 to 74 % a er the acid hydrolysis of the extracts.A similar trend was observed against the O 2 ˙-radical, which increased from 35.6 to 55.1 % a er the alkaline hydrolysis of the extracts.In this sense, the main contribution of this work was to demon-strate that the hydrolysis of extracts from vegetable matrices can considerably enhance their antioxidant activity.

Antimicrobial activity of crude and hydrolysed extracts of Brassica oleracea var. capitata
The antimicrobial activity of the crude and hydrolysed extracts (methanolic and ethanolic) was determined against Gram-positive Staphylococcus aureus and Gram--negative Escherichia coli.Table 4 presents the experimentally determined antimicrobial activities of the extracts obtained in this work.Only the hydrolysed extracts showed antimicrobial activity, where the alkaline hydrolysis extracts showed activity against Escherichia coli, whereas only the acid-hydrolysed methanolic extract had the activity against Staphylococcus aureus.However, the inhibitory halo cannot be considered as a confi rmation of antimicrobial activity unless its diameter is ≥8 mm (35).
The results obtained in this study are in good agreement with the literature; Gram-positive bacteria are more resistant than Gram-negative ones because they have thicker cell walls, ranging from 20-80 nm, whereas Gram--negative bacteria have a maximum wall thickness of 20 nm (36).Other authors have also evaluated the antimicrobial activity of Brassica oleracea extracts.Ayaz et al. (37) evaluated the antimicrobial activity of methanolic extracts of Brassica oleracea L. var.acephala using the same antimicrobial procedure and microorganisms as employed in this work, and verifi ed that the extracts had be er activity against S. aureus than against E. coli.
As verifi ed from the antioxidant activities of the extracts, the hydrolysed extracts of Brassica oleracea var.capitata had higher antimicrobial activities than the crude extracts.This result can be a ributed to a diff erence in the chemical profi les of the samples of the crude and hydrolysed extracts, as shown in Table 2.For example, alkaline hydrolysis enabled the identifi cation of hexadecanoic acid, which shows antimicrobial activity (30), justifying the highest antimicrobial activity of this extract.Another reason for be er performance of the hydrolysed extracts is that the extraction procedure used in this work resulted in the extraction of many glycosidic compounds, which have greater activity than the crude extracts a er hydrolysis.

Conclusions
This work presents important contribution to studying antioxidant activity using a combination of diff erent extraction conditions which allow the identifi cation of diff erent compounds from Brassica oleracea var.capitata.Specifi cally, the optimization of the ultrasound-assisted extraction of Bras sica oleracea var.capitata, the chemical characterization as well as the determination of antioxidant and antimicrobial activities of crude and hydrolysed extracts were successfully done.All extracts showed antioxidant activity against DPPH, O 2 ˙-and ROO˙ radicals; however, the hydrolysed extracts had considerably greater antioxidant activity.

Fig. 1 .
Fig. 1.Comparison between the presence and absence of ultrasound (US) in Brassica extraction a er 120 min (right Y-axis) and the contribution/increase of the US (le Y-axis) ). Li et al. (17) evaluated ultrasound-assisted extraction of carotenoids, Pingret et al. (18) investigated ultrasound-assisted water extraction of polyphenols from apple pomace, while Achat et al. (19) evaluated the enrichment of olive oil with oleuropein by ultrasound-assisted maceration.
The microorganisms used in the tests (Staphylococcus aureus ATCC 29213 and Escherichia coli ATCC 14948) were obtained from the Laboratory of Immunobiology and Molecular Glycobiology, Federal University of Viçosa, Viçosa, Brazil.Samples Brassica oleracea var.capitata (white cabbage) was bought in a local supermarket in Santa Maria, Rio Grande do Sul, Brazil, in January 2011.The sample was dried in an oven with air circulation at 60 °C for 72 h, according to Tanongkankit et al. (20).Next, the material was ground and stored at room temperature under a nitrogen atmosphere prior to extraction.

Table 1 .
Experimental conditions used to compare HPLC peak areas a er 120 min in the presence and absence of ultrasound

Table 2 .
Chemical profi le of crude and hydrolysed extracts of Brassica oleracea var.capitata obtained by ultrasound-assisted extraction

Table 3 .
Antioxidant activities of crude and hydrolysed extracts of Brassica oleracea var.capitata obtained by ultrasound-assisted extraction