ersity

Received in revised form

Keywords:Soybean

aracpro

solvent mixture composed of water, ethanol and ethyl acetate (40:40:20). Soybean meals presented 43%

soybe

contains high contents of bioactive compounds, mainly isoavones(Kao & Chen, 2006; Kao, Chien, & Chen, 2008) and soyasaponins(Wang, Wang, Lu, Kao, & Chen, 2009). These bioactive compoundsfrom soybeans and its products are associated with lower preva-lence of a number of chronic diseases, including cardiovascular

overlooked, eveny occur in higherBioactivity studiesand soyasaponinshumans should

g the whole inter-and soyasaponinssure to high tem-

peratures, oxygen and UV-light (Georgetti et al., 2008; Rostagno,Palma, & Barroso, 2005). However, to the best of our knowledge,the stability of bioactive compounds in soybean meal extracts hasnot been yet investigated.

Therefore, the aim of this work was to chemically characterizesoybean meal samples, with an emphasis on isoavones andsoyasaponins, in order to promote its use as a food ingredient andthus widen high added-value applications of this residue. Addi-tionally, we aimed to investigate, for the rst time, the stability of

* Corresponding author. Tel.: 55 21 3938 8213.E-mail addresses: silvafo@live.com (F.O. Silva), danielperrone@iq.ufrj.br

Contents lists availab

LWT - Food Science

w.e

LWT - Food Science and Technology 63 (2015) 992e1000(D. Perrone).grade oil for frying and cooking. The co-product from this processis a high-protein meal, known as soybean meal, mainly used asanimal feed. It may also be used for the production of soy proteinisolates and concentrates, which in turn are used as ingredients inmeat and bakery products, beverages, soups, infant formulas andother food products (Aguiar et al. 2012), thus aggregating economicvalue to this residue.

In addition to its content of high-quality protein, soybean meal

et al., 2009), while soyasaponins are usuallythough this class of bioactive compounds maamounts than isoavones (Kao & Chen, 2006).investigating the chronic effects of isoavonesisolated from soybean meal in animals andconsider the stability of these compounds durinvention period, since it is known that isoavonescontents and/or prole may be affected by expowidely consumed in Western countries, where soybean is mostcommonly crushed and the oil is extracted and rened into food-

compounds for metabolic and bioactivity studies. However, papershave focused on soybean meal as a source of isoavones (WangSoybean mealIsoavonesSoyasaponinsStability

1. Introduction

Differently from Asian countries,http://dx.doi.org/10.1016/j.lwt.2015.04.0320023-6438/ 2015 Elsevier Ltd. All rights reserved.higher protein content, from 29% to 101% higher bioactive compounds contents and 52% higher anti-oxidant capacity than soybeans. High moisture thermal procedure employed during soybean mealprocessing led to a 13-fold increase in aglycone isoavones contents, which could affect the bioavail-ability of isoavones in this residue. In addition to a 29-fold higher extraction yield, bioactive compoundsshowed higher stability in ternary solvent mixture extracts in comparison to ethanol, independently ofthe sample or storage conditions. We concluded that dry soybean meal extracts are suitable materials forperforming long-term in vivo studies, as these extracts were stable when stored at room temperatureunprotected from light for 180 days.

2015 Elsevier Ltd. All rights reserved.

an-based foods are not

diseases and certain types of cancer (Georgetti, Casagrande, Souza,Oliveira, & Fonseca,, 2008).

Soybean meal may be used as a source of various bioactiveAccepted 13 April 2015Available online 23 April 201512 April 2015future metabolic and bioactivity in vivo studies, we conducted, for the rst time, a 6-month stabilitystudy of soybean meal dry extracts. Soybean meal extracts were obtained either by ethanol or a ternaryCharacterization and stability of bioactivmeal

Fabricio de Oliveira Silva, Daniel Perrone*

Laboratorio de Bioqumica Nutricional e de Alimentos, Chemistry Institute, Federal Univesala 528A, 21941-909, Rio de Janeiro, Brazil

a r t i c l e i n f o

Article history:Received 22 December 2014

a b s t r a c t

In the present study we chof soybean oil extraction,

journal homepage: wwcompounds from soybean

of Rio de Janeiro, Av. Athos da Silveira Ramos 149, CT, Bloco A,

terized soybeans and soybean meals, which is the high protein co-productduced at industrial or laboratory scale. Additionally, for the purpose of

le at ScienceDirect

and Technology

lsevier .com/locate/ lwt

2,4,6-tris(2-pyridyl)-S-triazine (TPTZ), potassium persulfate, uo-

(Gardena, CA). Iron (II) sulfate was purchased from Merck (Darm-

1690 g, the supernatant was collected and the residue was re-

enceextracted twice as described above. The extraction procedure foreach sample was conducted in duplicate. These extracts wereanalyzed for the determination of phenolic compounds, avonoidsand saponins by spectrophotometric assays, and isoavones andin a domestic processor and dried for 5 h at 60 C in an oven. Afterdrying, the sample was additionally ground in an analytical mill(IKA A11 basic S1) in order to maximize oil extraction. 20 g ofsample was placed in a cellulose cartridge (33 94 mm) and theextraction procedure was conducted in a Soxhlet extractor for 7 husing n-hexane. The meal was left in a fume hood until dry andstored at room temperature.

2.3. Proximate composition

Moisture, ash, proteins and lipids were determined by the of-cial methods of the AOAC (2000). Total carbohydrates were deter-mined by difference. Each sample was analyzed in triplicate.

2.4. Bioactive compounds analysis

Soybean and soybean meal samples were extracted with aternary solvent mixture (water:ethanol:ethyl acetate e 40:40:20).Briey, 2 g of sample were extracted with 10 mL of the ternarymixture in a vortex for 2 min. After centrifugation for 10 min atstadt, Germany). Daidzin, glycitin, genistin, daidzein, glycitein,genistein, soyasapogenol B and soyasaponins B-I, B-II and B-IIIstandards were purchased from Apin Chemicals (Abingdon, UK).Saponin standard from soybean was purchased from Wako PureChemical Industries (Osaka, JP). All solvents were HPLC grade fromTedia (Faireld, OH). Milli-Q water (Millipore, Bedford, MA) wasused throughout the experiments.

2.2. Samples

Two sets of samples comprised of soybeans and soybean mealwere investigated. In both sets, samples were from the same batch.One set was donated by a soybean oil crush industry, while theother was produced in our laboratory using commercial soybeanspurchased in local markets.

Lab-scale soybean meal production was analog to industrialprocessing, with adaptations. Seeds were cracked and the hullsseparated using an air stream. Dehulled soybeans were wet andincubated at 60 C for 1 h in an oven. The sample was occasionallywater-sprayed to keep the beans wet. The seeds were then groundrescein, 2,20-azobis(2-methylpropionamidine) dihydrochloride(AAPH) and ()-6-hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid (Trolox) were purchased from SigmaeAldrich (St.Louis, MO). Sodium carbonate, vanillin and aluminum chloridewere purchased from Spectrum Chemical Manufacturing Corp.isoavones and soyasaponins in dry extracts obtained from soy-bean meal under different storage conditions, for future use inlong-term bioactivity studies.

2. Materials and methods

2.1. Standards and chemicals

Folin-Ciocalteau reagent, gallic acid, 2,20-azino-bis-(2-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS),

F.O. Silva, D. Perrone / LWT - Food Scisoyasaponins by LC-DAD-MS.2.4.1. Spectrophotometric assaysTotal phenolic content was determined by the Folin-Ciocalteau

reagent assay, as described by Singleton, Orthofer, and Lamuela-Raventos (1999). Quantication was performed using a gallic acidcalibration curve. Results were expressed as mg of gallic acidequivalents per g on a dry weight basis (dwb) (mg GAE/g).

Total avonoid content was determined by the spectrophoto-metric assay described by Taie, El-Mergawi, and Radwan (2008).Quantication was performed using a genistein calibration curve.Results were expressed as mg of genistein equivalents per g dwb(mg GE/g).

Total saponin content was determined by the vanillin-sulfuricacid assay, as described by Shiau et al. (2009). Quantication wasperformed using a calibration curve prepared with a commercialmixture of saponins obtained from soybeans. Results wereexpressed as mg of total saponins per g dwb.

The antioxidant capacity of the extracts was determined byFRAP (Ferric Reducing Ability of Plasma) and TEAC (Trolox Equiv-alent Antioxidant Capacity) assays. The FRAP assay was performedaccording to Benzie and Strain (1996) with slight modications. TheTEAC assay was performed according to Re, Pellegrini, andProteggente (1999) with slight modications.

Each sample was independently analyzed by each assay intriplicate.

2.4.2. Isoavones and soyasaponinsIsoavones and soyasaponins were simultaneously analyzed by

LC-DAD-MS as described by Fonseca, Villar, Donangelo, and Perrone(2014). The LC system (Shimadzu, Kyoto, Japan) comprised an LC-10ADvp quaternary pump, a CTO-10ASvp column oven, an 8125manual injector (Rheodyne) with a 20 mL loop and an SPD-M10Avpdiode array detector (DAD). This LC system was coupled to anLCeMS 2010 mass spectrometer (MS) (Shimadzu, Kyoto, Japan)equipped with an electrospray ion source. Chromatographic sepa-ration was achieved using a Kromasil C18 column (150 2.1 mm,5 mm,100 ) maintained at a constant temperature of 40 C. The LCtwo-phase mobile system consisted of a gradient of water (eluentA) and acetonitrile (eluent B), both added with 0.3 g/100 mL formicacid, with a constant ow rate of 0.3 mL/min. Prior to injection, thecolumnwas equilibratedwith 15% B. After injection, this proportionwas modied to 23% B in 1 min, kept constant until 23 min andincreased to 50% B until the end of the 35 min run. Twenty minintervals were used to re-equilibrate the column with 15% B. Ex-tracts were ltered through a 0.45 mm PTFE lter unit.

Isoavones and soyasaponins were monitored by DAD between190 and 370 nm and by MS using positive ionization, with anebulizer gas (N2) ow of 3.0 L/min, operated in the single ionmonitoring (SIM) mode. Identication of aglycone and glycosylatedisoavones was performed by comparison with retention time andpseudomolecular ion of the respective standard. Identication ofsoyasaponins was performed by comparison with retention timeand the most abundant ion of the respective standard ([MH] forsoyasaponin B-I and [M-sugar-H2O H] for both soyasaponins B-II and B-III). Identication of compounds for which there were nocommercial standards available (malonylglucosilated and acetyl-glucosilated isoavones) was performed by the pseudomolecularion in the MS (Supplementary Figure 1AeE).

Quantication was performed by external standardization. Iso-avones were quantied by their DAD peak areas at 250 nm. Thecontents of malonylglucosilated and acetylglucosilated isoavoneswere determined from the calibration curve of the correspondingb-glucosylated isoavone. Soyasaponins (B-I, B-II and B-III) werequantied by their DAD peak areas at 195 nm. Soyasaponins B-IIand B-III were quantied together, as it was not possible to chro-

and Technology 63 (2015) 992e1000 993matographically separate these compounds. Data were acquired by

nceLCMS solution software (Shimadzu Corp., version 2.00, 2000). Re-sults were expressed as mg of compound per g dwb. Each samplewas analyzed in duplicate.

This chromatographic method was previously validated(Fonseca et al., 2014) and was shown to be linear (R2 > 0.994 for allanalytes), reproducible (inter-day precision of 10.9%) and presentedsatisfactory limits of quantication for both isoavones (

Fig. 1. Proximate composition (A), contents of total phenolics (TP), avonoids (TF) and saponins (TS) (B), and antioxidant capacity (C) in commercial soybean ( ), experimental meal( ), industry soybean ( ) and industry meal ( ) samples. Values expressed as mean standard deviation (n 3); different superscript letters indicates signicant difference(ANOVA, Tukey's post-test; p < 0.05); GAE gallic acid equivalents; GE genistein equivalents; TE Trolox equivalents.

Table 1Isoavones and soyasaponins contents (mg/g dwb) in soybean and soybean meal experimental and industrial samples.a

Compound MS identication Experimental Industrial

m/z Ion Soybean Soybean meal Soybean Soybean meal

Isoavonesb-glucosylated formsDaidzin 417 [MH] 0.34 0.03a 0.06 0.00c 0.17 0.01b 0.34 0.01aGlycitin 447 [MH] 0.07 0.01b 0.03 0.00c 0.08 0.00b 0.15 0.01aGenistin 433 [MH] 0.46 0.03a 0.09 0.00c 0.26 0.00b 0.53 0.02aSub total 0.87 0.08a 0.17 0.01c 0.51 0.01b 1.02 0.04aMalonyl glucosylated formsDaidzin 519 [MH] 0.28 0.02a 0.27 0.00a 0.27 0.01a 0.24 0.01aGlycitin 549 [MH] NDb ND ND NDGenistin 535 [MH] 0.03 0.00b 0.03 0.00b 0.05 0.00a 0.04 0.00aSub total 0.31 0.03a 0.30 0.00a 0.32 0.01a 0.28 0.02aAcetyl glucosylated formsDaidzin 459 [MH] 0.03 0.00b 0.03 0.00b 0.02 0.00b 0.05 0.00aGlycitin 489 [MH] ND ND ND 0.03 0.00aGenistin 475 [MH] 0.03 0.00b 0.01 0.00b 0.00 0.00b 0.21 0.02aSub total 0.06 0.01b 0.03 0.00b 0.03 0.00b 0.28 0.02aAglyconesDaidzein 255 [MH] 0.04 0.00b 0.36 0.01a 0.05 0.00b 0.09 0.01cGlycitein 285 [MH] 0.01 0.00d 0.36 0.01a 0.05 0.00c 0.09 0.00bGenistein 271 [MH] 0.02 0.00c 0.37 0.02a 0.04 0.00c 0.08 0.00bSub total 0.07 0.01c 1.09 0.03a 0.14 0.00c 0.26 0.01bTotal 1.31 0.12b 1.60 0.04a 1.00 0.03b 1.84 0.08aSoyasaponinsB-I 944 [MH] 1.13 0.06b 2.51 0.11a 1.42 0.10b 2.95 0.16aB-II BIII 423 [M-sugar-H2O H] 0.46 0.03d 0.78 0.00b 0.55 0.01c 0.91 0.00aTotal 1.60 0.03c 3.29 0.11b 1.96 0.11c 3.86 0.16a

a Values expressed as mean standard deviation (n 2); means in the same rowwith different online letters are signicantly different (ANOVA, Tukey's post-test; p < 0.05).b Not detected.

F.O. Silva, D. Perrone / LWT - Food Science and Technology 63 (2015) 992e1000 995

g) using a ternary mixture of methanol, acetonitrile and water assolvent. Kao et al. (2008) also reported higher total isoavonescontents (up to 3.3 mg/g) than those found in the present study fora soybean meal sample subjected to supercritical carbon dioxideextraction at different pressures and temperatures.

the relative content of malonylglucoside isoavones and a 5.3-foldincrease in the relative content of acetylglucoside forms, theexperimental conditions used in our laboratory led to a drasticreduction (84%) of b-glucoside isoavones relative content and a12.9-fold increase in the relative content of the aglycone forms. Ithas already been reported that isoavones prole changes greatlydepends on thermal processes employed during food preparation,including cooking, baking, oven-drying and roasting (Georgettiet al., 2008). Low moisture thermal processing was reported tocause the conversion of malonyl to acetylglucoside isoavones.Therefore, soybeanmeal roasting usually applied after industrial oilextraction (Pananun, Montalbo-Lomboy, Noomhorm, Grewell, &Lamsal, 2012) probably explains the isoavone prole changeobserved in the industrial soybean meal sample. In contrast, highmoisture thermal processing leads to hydrolysis of b-glucosidesisoavones to aglycones, probably due to an increase in b-glycosi-dase activity (Lima & Ida, 2014). This phenomenon may explain theobserved shift in isoavone prole during the production of soy-bean meal under laboratory conditions, as soybeans wereconstantly water-sprayed during drying prior to oil extraction.

The experimental soybean meal sample, which is the richest inaglycone forms (1.09 mg/g) (Table 1), might be a more bioavailablesource of isoavones because these forms are absorbed faster andin greater amounts than glucosides in human gastrointestinal tract(Aguiar et al., 2012; Okabe, Shimazu, & Tanimoto, 2011). Eventhough industrial processing increased total isoavones contents to

Fig. 2. Aglycone ( ), acetylglucosilated ( ), malonylglucosilated ( ), and b-glucosi-lated ( ) isoavones relative content from soybean and soybean meal experimentaland industrial samples.

F.O. Silva, D. Perrone / LWT - Food Science and Technology 63 (2015) 992e1000996Despite the equivalent total isoavones contents betweenexperimental and industrial soybean meals, LC-DAD-MS analysisrevealed different proles between these samples (Fig. 2). Indus-trial soybean meal sample presented a b-glucoside and aglyconeisoavones prole similar to that of the corresponding soybean,representing 53% and 14%, on average, of total isoavones,respectively. In contrast, experimental soybean meal sample pre-sented a malonyl and acetylglucoside isoavones prole similar tothat of the corresponding soybean, representing 20% and 3%, onaverage, of total isoavones, respectively. While industrial pro-cessing of soybeans for oil production resulted in a 48% decrease inFig. 3. Stability of bioactive compounds and antioxidant capacity (FRAP and TEAC) of dry expof dry industrial soybean meal extract obtained with ethanol ( ) or ternary mixture ( ) dunprotected from light [ ]) for 180 days.a higher extent, high moisture thermal processing, such as theprocedure employed in our laboratory, as well as fermentationprocesses of soybean meal (Okabe et al., 2011), could be used toimprove functional value of this co-product, broadening its use infood products with high isoavones bioavailability.

In the present study, we investigated the contents of three groupB soyasaponins and their respective aglycone, sapogenol B.Although the industrial and experimental soybean samples showedequivalent soyasaponins contents (on average of 1.78 mg/g), theindustrial soybean meal sample presented a 17% higher content ofthese compounds than the experimental one (Table 1). The major

erimental soybean meal extract obtained with ethanol ( ) or ternary mixture ( ) anduring storage at 20 C (freezer) ( ) and room temperature (protected [ ] and

enceF.O. Silva, D. Perrone / LWT - Food Scisoyasaponin in all samples was soyasaponin B-I, which corre-sponded to 72% of total soyasaponins in both soybean samples andto 76% in both soybean meal samples. Soyasaponins B-II and B-IIIaccounted together for the remainder of soyasaponins content.Soyasapogenol B was not found in any sample.

Berhow, Kong, Vermillion, and Duval (2006) reported totalsoyasaponin contents ranging from 4.73 to 5.75 mg/g for soybeansand from 5.86 to 6.58 mg/g for defatted soy ours. As observed inthe present study, themajor soyasaponin in all samples analyzed byBerhow et al. (2006) was soyasaponin I. Although the contentsreported by Berhow et al. (2006) were up to 2.2-fold higher thanthose found in the present study, it should be noted that thoseauthors analyzed 13 soyasaponins, including soyasaponin bg,which was the second most abundant soyasaponin in their

Fig. 4. Stability of isoavones and soyasaponins of dry experimental soybean meal extractextract obtained with ethanol ( ) or ternary mixture ( ) during storage at 20 C (freezer)for 180 days.and Technology 63 (2015) 992e1000 997samples. Considering only the contents of the soyasaponinsanalyzed in the present study (B-I, B-II and B-III), the contentsfound in our samples were up to 79% higher than those reported byBerhow et al. (2006). Unfortunately, we could not analyze soyasa-ponin bg in our samples due to the lack of a commercial standard.

3.4. Bioactive compounds stability

Stability of bioactive compounds and antioxidant capacityevaluated by spectrophotometric assays during storage of extractsat 20 C (freezer) and at room temperature (protected and un-protected from light) are depicted in Fig. 3. Stability of isoavonesand soyasaponins analyzed by LC-DAD-MS at the same conditionsare depicted in Fig. 4. Ternary mixture extraction yield (on average

obtained with ethanol ( ) or ternary mixture ( ) and of dry industrial soybean meal( ) and room temperature (protected [ ] and unprotected from light [ ])

Table 2Contents of bioactive compounds and antioxidant capacity at t 0 and t 180 days in dry soybeanmeal (experimental and industrial) extracts (obtainedwith ethanol or ternary solventmixture) during storage at20 C (freezer)and room temperature (protected and unprotected from light).a

Component Experimental Industrial

Ethanol Ternary Ethanol Ternary

t 0 t 180 days t 0 t 180 days t 0 t 180 days t 0 t 180 daysRTlight RTdark Freezer RTlight RTdark Freezer RTlight RTdark Freezer RTlight RTdark Freezer

Total phenolicsb

(mg GAE/g)11.1 10.3 11.1 10.3 12.4 15.9* (28%) 17.7* (43%) 13.1 4.8 4.4 4.7 4.9 10.5 13.2* (26%) 12.9* (23%) 12.0* (15%)

Total avonoidsb

(mg GE/g)13.6 10.3* (24%) 10.4* (24%) 8.6* (36%) 7.3 8.0* (11%) 9.4* (30%) 6.7 11.0 5.8* (47%) 7.5* (32%) 7.3* (34%) 10.0 11.0* (10%) 10.8* (8%) 10.1

Total saponinsb

(mg/g)71.5 52.5* (30%) 50.9* (32%) 71.4 225.0 190.9* (15%) 230.8 172.2* (23%) 43.9 29.6 27.3* (38%) 36.7 253.1 234.6 216.4* (14%) 200.4* (21%)

FRAPb

(mmol Fe2/g)74.7 35.1* (53%) 36.2* (52%) 32.8* (56%) 71.4 92.4* (30%) 104.8* (47%) 78.5* (10%) 48.0 34.1* (29%) 42.4* (10%) 30.6* (36%) 91.6 113.0* (23%) 111.9* (22%) 99.9* (9%)

TEACb

(mmol Trolox/g)103.1 39.8* (61%) 49.4* (52%) 62.3* (40%) 105.0 49.8* (53%) 95.6 86.9 71.7 24.4* (66%) 11.3* (84%) 39.4* (45%) 109.5 107.5 78.03* (29%) 117.0

Total isoavonesc

(mg/g)10.4 8.6 11.0 10.0 5.3 7.3* (38%) 9.2* (76%) 6.1 6.2 5.7 5.9 6.3 5.6 8.9* (59%) 8.7* (57%) 8.1* (46%)

b-glucosidesc

(mg/g)1.0 1.2* (20%) 1.3* (33%) 1.1 0.7 1.1* (49%) 1.5* (106%) 0.9* (29%) 3.9 3.9 4.1 4.1 3.6 6.1* (67%) 6.0* (64%) 5.5* (53%)

Malonylglucosidesc

(mg/g)0.6 0.4* (33%) 0.5* (21%) 0.6 1.1 1.3* (13%) 1.6* (46%) 1.5* (30%) 0.3 0.1* (50%) 0.2* (24%) 0.3 0.6 0.9* (41%) 0.8* (33%) 0.8* (39%)

Acetylglucosidesc

(mg/g)0.1 0.09 0.1 0.1 0.1 0.1 0.2* (90%) 0.1 0.3 0.3 0.3 0.3 0.4 0.6* (43%) 0.6* (43%) 0.5* (33%)

Aglyconesc (mg/g) 8.6 6.9 9.1 8.1 3.3 4.7* (44%) 5.8* (78%) 4.8* (47%) 1.6 1.3 1.3 1.5 0.8 1.3* (48%) 1.3* (48%) 1.2* (47%)Soyasaponinsc

(mg/g)12.7 7.5* (41%) 7.9* (38%) 7.6* (41%) 13.2 13.9 18.3* (39%) 15.5 9.0 7.1 6.9 6.1* (32%) 15.3 20.4* (33%) 22.0* (44%) 18.9* (24%)

a Means at t 180 days with a superscript asterisk are signicantly different from their corresponding sample at t 0 (ANOVA, Tukey's post-test; p < 0.05). Relative change in relation to t 0 is shown in parenthesis.GAE gallic acid equivalents; GE genistein equivalents.

b Spectrophotometric analysis.c LC-DAD-MS analysis.

F.O.Silva,D

.Perrone/LW

T-Food

Scienceand

Technology63

(2015)992

e1000

998

increase in soyasaponins B-I, B-II and B-III contents in extracts

raised to explain the increase of isoavones in these extracts could

solvent, due to higher energy consumption for solvent recycling.

ethanol and ethyl acetate (40:40:20) as this solvent led to higher

ence21.0%) was much higher than that obtained with ethanol (onaverage 1.4%). Moreover, extraction solvent was the parameter withthemost prominent effect on bioactive compounds and antioxidantcapacity stability. In general, ethanolic extracts showed lower sta-bility than those obtained with the ternary mixture, independentlyof the sample (experimental or industrial) or the storageconditions.

While TP, TF and FRAP increased after 180 days of storage (8%e47%) in extracts obtained with the ternary mixture, correspondingethanolic extracts presented a constant behavior for TP and adecrease in TF and FRAP (10%e56%) (Table 2). Although TS and TEACwere unstable for most samples and conditions, a higher decrease(from 30% to 84%) was observed for ethanolic extracts in compar-ison to those obtained with the ternary mixture (from 14% to 53%)(Table 2).

Total isoavones behaved similarly to TP, showing stability inethanolic extracts regardless of the sample and storage conditions.Ethanolic extracts obtained from the experimental soybean mealsample stored at room temperature showed an average decrease of27% in malonylglucoside isoavones accompanied by an identicalaverage increase in b-glucoside isoavones. Ethanolic extracts ob-tained from the industrial soybean meal sample stored in the sameconditions also showed a decrease in malonylglucoside isoavones(on average 37%) (Table 2), but the corresponding increase in b-glucoside forms was only observed until 120 days and not at theend of the storage period. Similarly, Rostagno et al. (2005) reportedthat during short term storage (7 days) of dry extracts obtainedwith ethanol:water (50:50) malonyl forms were the most suscep-tible to degradation, yielding b-glucoside forms. In our study, iso-avones contents and prole of all ethanolic extracts storedat 20 C remained unchanged during the whole storage period(Table 2), corroborating with Rostagno et al. (2005). These authorsconcluded that temperature was the single most important factorto affect isoavones stability and suggested that extracts should bekept at temperatures lower than 10 C and protected from light toavoid degradation. Nevertheless, Xu, Wu, and Godber (2002) re-ported that b-glucoside forms could resist to heat-induceddecomposition at temperatures near 100 C in the absence ofother factors, such as enzymes and only became unstable whenheating temperatures were increased above 135 C.

Surprisingly, total isoavones contents increased (38%e76%) inextracts obtained with the ternary mixture after 180 days of stor-age. This increase, however, did not signicantly modify the iso-avone prole of these extracts, as the concentration of allisoavones forms increased similarly (from 13% to 90%) (Table 2). Itis known that isoavones may establish different types of in-teractions with soy proteins, especially glycinin and b-conglycinin,because of their diverse polarity and hydrophobicity as well as theirability to form hydrogen bonds (Speroni, Milesi, & A~non, 2010).During long-term storage, structure of soy proteins such as glycininmight change (Saio, Kobayakawa,& Kito,1982), due to a decrease infree sulfhydryl groups and an increase in disulfhydryl groups.Moreover, the hydrogen bondings within the glycinin moleculemight decrease because of an increase in b-helix structure and adecrease in the b-sheet structure (Hou & Chang, 2004). Thesestructural changes might have decreased the interaction of iso-avones with soy proteins during storage, therefore increasingtheir extractability. In addition, the aforementioned stability ofisoavones in ethanolic extracts (Table 2) strengthens this hy-pothesis and might be explained by the solubility of soy proteins inthis solvent, which is lower than in the ternary mixture, whichcontains 40% water (Pace, Trevi~no, Prabhakaran, & Scholtz, 2004).In fact, protein contents in the extracts obtained with the ternarymixture (7.62 g/100 g) were 1.8 times higher than those of ethanolic

F.O. Silva, D. Perrone / LWT - Food Sciextracts (4.17 g/100 g).extraction yields and stability of bioactive compounds. For thispurpose, these extracts may be stored at room temperature un-protected from light for at least 180 days.

Acknowledgments

The authors gratefully thanks CAPES (Brazil) for nancial sup-port and Mr. Silva's scholarship and UFRJ, CNPq (461464/2014-4)and FAPERJ (Brazil) (E-26/110.420/2012 and E-26/102.974/2012) fornancial support.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.lwt.2015.04.032.

References

Aguiar, C. L., Haddad, R., Eberlin, M. N., Carrao-Panizzi, M. C., Mui, T. S., & Park, Y. K.(2012). Thermal behavior of malonylglucoside isoavones in soybean ouranalyzed by RPHPLC/DAD and eletrospray ionization mass spectrometry. LWT eFood Science and Technology, 48(1), 114e119.

AOAC (Association of Ofcial Analytical Chemists. (2000). Ofcial methods of analysis(17th ed.). Gaithersburg, MD, USA: Association of Analytical Communitie.

Benzie, I. F., & Strain, J. J. (1996). The ferric reducing ability of plasma (FRAP) as aNevertheless, the use of the ternary mixture is feasible whenconsidering the small-scale production of extracts intended forbiological studies.

4. Conclusion

Our results suggest that soybean meals would be more inter-esting as an ingredient than soybeans themselves for the devel-opment of functional foods, since the former showed higherprotein and bioactive compounds contents as well as higher anti-oxidant capacity than the latter. Moreover, high moisture thermalprocessing of soybeans is a procedure that could be employed toimprove the functional value of the obtained co-product as it seemsto yield meals with potentially higher isoavones bioavailability.

Soybean meal extracts for long-term bioactivity studies shouldbe preferably obtained using a ternary mixture composed of water,also explain the increase in soyasaponins, as these compounds alsoform very tight linkages to soy proteins (Potter, Jimenez-Flores,Pollack, Lone, & Berber-Jimenez, 1993).

Although the ternary mixture led to higher extraction yields andlonger stability of bioactive compounds, it may not be practical,from an industrial point of view, to produce extracts using thisobtained with the ternary mixture may be related to the degrada-tion of the corresponding DDMP conjugated soyasaponins, bg, baand gg, respectively. The conversion of soyasaponin bg to soyasa-ponin B-I during storage and extraction has been reported in peas(Daveby, man, Betz, &Musser, 1998). Additionally, the hypothesisIn general, soyasaponins contents increased (24%e39%) in ex-tracts obtained with the ternarymixture, whereas decreased (32%e41%) in ethanolic extracts. Although these compounds are generallyconsidered stable at room temperature for short time periods (up to60 min) (Heng et al., 2006), saponins from Aspargus racemosus drypowder stored for 12 months at room temperature were highlydegraded (up to 50%) (Madan, Yadav, & Tyagi, 2005). The observed

and Technology 63 (2015) 992e1000 999measure of antioxidant power: the FRAP assay. Analytical Biochemistry, 239(1),70e76.

Berhow, M. A., Kong, S. B., Vermillion, K. E., & Duval, S. M. (2006). Completequantication of group A and group B soyasaponins in soybeans. Journal ofAgricultural and Food Chemistry, 54(6), 2035e2044.

Bolanho, B., & Beleia, A. (2011). Bioactive compounds and antioxidant potential ofsoy products. Alimentos e Nutri~ao Araraquara, 22(4), 539e546.

Daveby, Y. D., man, P., Betz, J. M., & Musser, S. M. (1998). Effect of storage andextraction on ratio of soyasaponin I to 2,3-dihydro-2,5-dihydroxy-6-methyl-4-pyrone-conjugated soyasaponin I in de-Hulled peas (Pisum sativum L.). Journalof the Science of Food and Agriculture, 78(1), 141e146.

Fonseca, N. D., Villar, M. P. M., Donangelo, C. M., & Perrone, D. (2014). Isoavonesand soyasaponins in soy infant formulas in Brazil: prole and estimated con-sumption. Food Chemistry, 143, 492e498.

Georgetti, S. R., Casagrande, R., Souza, C. R. F., Oliveira, W. P., & Fonseca, M. J. V.(2008). Spray drying of the soybean extract: effects on chemical properties andantioxidant activity. LWT e Food Science and Technology, 41(8), 1521e1527.

Grieshop, C. M., Kadzere, C. T., Clapper, G. M., Flickinger, E. A., Bauer, L. L.,Frazier, R. L., et al. (2003). Chemical and nutritional characteristics of UnitedStates soybeans and soybean meals. Journal of Agricultural and Food Chemistry,51(26), 7684e7691.

Heng, L., Vincken, J. P., Hoppe, K., van Konigsveld, G. A., Decroos, K., Gruppen, H.,et al. (2006). Stability of pea DDMP saponin and the mechanism of itsdecomposition. Food Chemistry, 99(2), 326e334.

Hou, D., & Chang, S. (2004). Structural characteristics of puried glycinin fromsoybeans stored under various conditions. Journal of Agricultural and FoodChemistry, 3792e3800.

Joshi, A. U., Liu, C., & Sathe, S. L. (2015). Functional properties of select seed ours.LWT e Food Science and Technology, 60(1), 325e331.

Kao, T.-H., & Chen, B. (2002). An improved method for determination of isoavonesin soybean powder by liquid chromatography. Chromatographia, 423(7),423e430.

Kao, T.-H., & Chen, B.-H. (2006). Functional components in soybean meal and theireffects on antioxidant activity. Journal of Agricultural and Food Chemistry, 54(20),

Pananun, T., Montalbo-Lomboy, M., Noomhorm, A., Grewell, D., & Lamsal, B. (2012).High-power ultrasonication-assisted extraction of soybean isoavones and ef-fect of toasting. LWT e Food Science and Technology, 47(1), 199e207.

Potter, S. M., Jimenez-Flores, R., Pollack, J., Lone, T. A., & Berber-Jimenez, M. D.(1993). Protein-saponin interaction and its inuence on blood lipids. Journal ofAgricultural and Food Chemistry, 41, 1287e1291.

Re, R., Pellegrini, N., & Proteggente, A. (1999). Antioxidant activity applying animproved ABTS radical cation decolorization assay. Free Radical Biology &Medicine, 26(9/10), 1231e1237.

Rostagno, M. A., Palma, M., & Barroso, C. G. (2005). Short-term stability of soyisoavones extracts: sample conservation aspects. Food Chemistry, 93(3),557e564.

Saio, K., Kobayakawa, K., & Kito, M. (1982). Protein denaturation during modelstorage studies of soybeans and meals. Cereal Chemistry, 59(5), 408e412.

Sakthivelu, G., Akitha Devi, M. K., Giridhar, P., Rajasekaran, T., Ravishankar, G. A.,Nikolova, M. T., et al. (2008). Isoavone composition, phenol content, andantioxidant activity of soybean seeds from India and Bulgaria. Journal of Agri-cultural and Food Chemistry, 56(6), 2090e2095.

Shiau, I.eL., Shih, T.-L., Wang, Y.-N., Chen, H.-T., Lan, H.-F., Lin, H.-C., et al. (2009).Quantication for saponin from a soapberry (Sapindus mukorossi Gaertn) incleaning products by a chromatographic and two colorimetric assays. Journal ofthe Faculty of Agriculture Kyushu University, 54(1), 215e221.

Singleton, V. L., Orthofer, R., & Lamuela-Raventos, R. M. (1999). Analysis of totalphenols and other oxidation substrates and antioxidants by means of folin-ciocalteu reagent. Methods in Enzymology, 299, 152e178.

Speroni, F., Milesi, V., & A~non, M. C. (2010). Interactions between isoavones andsoybean proteins: applications in soybean-proteineisolate production. LWT eFood Science and Technology, 43(8), 1265e1270.

Taie, H. A. A., El-Mergawi, R., & Radwan, S. (2008). Isoavonoids, avonoids,phenolic acids proles and antioxidant activity of soybean seeds as affected byorganic and bioorganic fertilization. Journal of Agricultural and EnvironmentalSciences, 4(2), 207e213.

Wang, H., & Murphy, P. A. (1994). Isoavone content in commercial soybean foods.Journal of Agricultural and Food Chemistry, 42(8), 1666e1673.

F.O. Silva, D. Perrone / LWT - Food Science and Technology 63 (2015) 992e100010007544e7555.Kao, T.-H., Chien, J.-T., & Chen, B.-H. (2008). Extraction yield of isoavones from

soybean meal as affected by solvent and supercritical carbon dioxide. FoodChemistry, 107(4), 1728e1736.

Lima, F. S., & Ida, E. I. (2014). Optimisation of soybean hydrothermal treatment forthe conversion of b-glucoside isoavones to aglycones. LWT e Food Science andTechnology, 56(2), 232e239.

Madan, V. K., Yadav, O. P., & Tyagi, C. S. (2005). Post harvest degradation of saponincontent in powder of Asparagus racemosus tubers with different drying andstorage methods. Acta Horticulturae, 776, 261e268.

Malencic, D., Popovic, M., & Miladinovic, J. (2007). Phenolic content and antioxidantproperties of soybean (Glycine max (L.) Merr.) seeds. Molecules, 12, 576e581.

Okabe, Y., Shimazu, T., & Tanimoto, H. (2011). Higher bioavailability of isoavonesafter a single ingestion of aglycone-rich fermented soybeans compared withglucoside-rich non-fermented soybeans in Japanese postmenopausal women.Journal of the Science of Food and Agriculture, 91(4), 658e663.

Pace, C. N., Trevi~no, S., Prabhakaran, E., & Scholtz, J. M. (2004). Protein structure,stability and solubility in water and other solvents. Philosophical Transactions ofthe Royal Society of London Series B Biological Sciences, 359(1448), 1225e1234.Wang, B. F., Wang, J. S., Lu, J. F., Kao, T.-H., & Chen, B. H. (2009). Antiproliferationeffect and mechanism of prostate cancer cell lines as affected by isoavonesfrom soybean meal. Journal of Agricultural and Food Chemistry, 57(6),2221e2232.

Wu, J., & Muir, A. D. (2010). Isoavone during protease hydrolysis of defatted soy-bean meal. Food Chemistry, 118(2), 328e332.

Xu, B., & Chang, S. K. C. (2008). Characterization of phenolic substances and anti-oxidant properties of food soybeans grown in the North Dakota-Minnesotaregion. Journal of Agricultural and Food Chemistry, 56(19), 9102e9113.

Xu, B., & Chang, S. K. C. (2011). Reduction of antiproliferative capacities, cell-basedantioxidant capacities and phytochemical contents of common beans andsoybeans upon thermal processing. Food Chemistry, 129(3), 974e981.

Xu, Z., Wu, Q., & Godber, J. (2002). Stabilities of daidzin, glycitin, genistin, andgeneration of derivatives during heating. Journal of Agricultural and FoodChemistry, 50, 7402e7406.

Characterization and stability of bioactive compounds from soybean meal1. Introduction2. Materials and methods2.1. Standards and chemicals2.2. Samples2.3. Proximate composition2.4. Bioactive compounds analysis2.4.1. Spectrophotometric assays2.4.2. Isoflavones and soyasaponins

2.5. Stability of bioactive compounds from soybean meal extracts2.6. Statistical analysis

3. Results and discussion3.1. Proximate composition3.2. Total phenolics, flavonoids, saponins and antioxidant activity3.3. Isoflavones and soyasaponins by LC-DAD-MS3.4. Bioactive compounds stability

4. ConclusionAcknowledgmentsAppendix A. Supplementary dataReferences