Analysis of the Phenolic Content and Antioxidant Activities of Soybean Extracts from Different Regions and Cultivars

Abstract

Phenolic compounds such as isoflavones contribute to the antioxidant properties of soybean. This study investigated differences in the phenolic content and antioxidant activity of ten soybean cultivars grown in two different locations in Korea. We observed a wide range of variation in total isoflavone and phenolic content by location and cultivar. The total isoflavone content of cultivars grown at Pyeongchang (sub-highland) was significantly higher than that in cultivars grown at Gangneung (lowland). In contrast, the total phenolic content of cultivars grown at Gangneung was greater than that of cultivars grown at Pyeongchang. The radical scavenging activity of DPPH was similar to that of the total phenolic content rather than that of the total isoflavone content. These results suggest that antioxidant activity of soybean was associated with phenolic compounds rather than isoflavones. To identify the individual antioxidant components, we used an on-line HPLC-$ABTS^+$-based assay system, ESI/MS, and NMR. The results showed that the strongest antioxidant activity was linked to epicatechin.

Introduction

Studies suggest that free radicals and reactive oxygen species such as superoxide radicals, hydroxyl radicals, and peroxyl radicals are harmful to human health and that they trigger various conditions, such as cancer, coronary heart diseases, inflammatory disorders, and arteriosclerosis; they are also linked to aging (Ames et al., 1993; Nanjo et al., 1996; Lee et al., 2007; Li et al., 2007; Wang et al., 2007). Antioxidants may help prevent oxidative damage in the human body (Ames et al., 1993; Cos et al., 2003). As a result, scientific interest has increased in the levels of antioxidant components in food.

Reports have proposed that the antioxiadant activity of soybean (Glycine max L. Merrill) seeds and foods is mainly affected by phenolic compounds such as isoflavone, flavonoids, and phenolic acids (Malencic et al., 2007; Riedl et al., 2007). Many studies have suggested that the antioxidant activity of soybean seeds and soybean foods was positively correlated with total phenolics and isoflavones (Kim et al., 2006; Malencic et al., 2008; Devi et al., 2009). Several studies exhibit that genetic variability affects seed phenolic content of soybean cultivars, and phenolic content variations can also be affected by environmental conditions (Malencic et al., 2007; Hoeck et al., 2000; Lee et al., 2003). A number of studies have recently focused on the bioactivity of phenolic compounds, especially isoflavone in terms of its antioxidant and anticancer properties (Lamartiniere, 2000; Mezei et al., 2003).

The majority of studies conducted thus far have focused on the antioxidant activity and phenolic compounds of soybean. Few studies have attempted to identify the major components that contribute to the antioxidant activity of soybean. As soybean contains many different types of phenolic compounds (Kim et al., 2004; Lee et al., 2008), it is feasible that the composition and content of the compounds will influence the level of antioxidant activity.

The antioxidant activity and phenolic composition of soybeans may differ according to their genotype or environmental condition. To investigate the beneficial condition for higher antioxidant content, it is important to examine the different soybean cultivars in different environment. Profiling soybean antioxidant activity, phenolic content and isoflavone content can help elucidate what is the major factor for antioxidant activity of soybean and lead to soy products with greater health benefits. In this study, we evaluated the isoflavone, total phenolic content and antioxidant activity of ten soybean cultivars produced from different regions to identify the major antioxidant components of Korean soybean cultivars.

 

Materials and Methods

Soybean seed and plant cultivation

Ten popular soybean cultivars were grown at two regions; Gangneung (GN) (lowland, 20 m altitude) and Pyeongchang (PC) (sub-highland, 600 m altitude) in Korea in 2008. The cultivars differed in seed weight, seed coat color, and cotyledon color (Table 1). The soil type was a silt clay loam at all regions.

Table 1.zGangneung: 37°N, 128°E, 20 m altitude. yPyeongchang: 37°N, 128°E, 600 m altitude. xmean ± standard error.

Each plot consisted of ten rows (5 m long and 0.6 m between rows). The experiment consisted of a completely randomized design with three replicates. The planting arrangement was 60 ㎝ × 15 ㎝ per plot, and the plants were thinned to make a uniform density 20 days after planting. The plots were hand weeded during June and July and appropriate pesticides were applied to control diseases and insects. Fertilizers were applied prior to plowing at the recommended rates of 8, 8 and 12 ㎏ per 1,000 ㎡ for N, P2O5 and K2O, respectively. All other soybean cultivation procedures followed the standard methods used in Korea (Ryu et al., 2001) Soybean seeds were harvested from each replicate at each region. The harvested seed samples were freeze dried, ground, and stored at -80℃ prior to analysis of phytochemical and antioxidant activity.

Chemicals

The three aglycon standards of isoflaovne (daidzein, genistein, and glycitein), gallic acid, 2,2-diphenyl-1-picrylhydrazyl (DPPH) and 2,2'-azinobis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) were purchased from Sigma-Aldrich Chemicals (Saint Louis, MO, USA). Dimethyl sulfoxide–d6 was purchased from Cambridge Isotope Laboratory. All HPLC solvents were purchased from Fisher Scientific (Pittsburgh, PA, USA). All of the chemicals and solvents were analytical grade or HPLC grade.

Isoflavone analysis

Following acid hydrolysis of the endogenous 12 isoflavones to their aglycon forms, the concentraions of daidzein, genistein, and glycitein (summed to obtain total isoflavone) were determined using a modified version of the HPLC method of Franke et al. (1994).

Soybean seeds were powdered using a mill with a ㎜ sieve. Two grams of dry ground sample were suspended in 10 ㎖ of 1 N HCl and heated for 2 h at 100℃. The extracts were allowed to cool, mixed with 30 ㎖ of methanol, and stirred for 24 h at 40℃. The solution was then filtered through Whatman No. 42 filter paper, and the filter was rinsed three times with methanol. The filtrate was transferred to a 50 ㎖ volumetric flask. The sample was filtered through a 0.45 ㎛ PTFE filter (Pall, Ann Arbor, MI, USA, and 1 ㎖ of the filtrate was transferred to sample vials for HPLC analysis.

The concentrations of individual hydrolyzed daidzein, genistein, and glycitein were determined using a HPLC system (Waters, Alliance system, Milford, MA, USA) consisting of a 2695 Separation Module pump and a 2996 Photodiode Array Detector. Twenty microliters of sample were injected into a Phenomenex ODS Hypersil C18 column (5 ㎛ particle size, 250 ㎜ × 4.6 ㎜, Phenomenex, Torrance, CA, USA). A linear HPLC gradient was used; solvent A was 0.1% aq. acetic acid in nanopure water, and solvent B was 0.1% aq. acetic acid in acetonitrile. Following the injection of the sample, solvent B was increased from 15 to 35% for 50 min and then maintained at 35% for 10 min. The solvent flow rate was 1 mL/min. The wavelength of the UV detector was set at 254 ㎚. The MassLynx software associated with the Waters HPLC instrument was used to generate linear calibration curves based on peak areas for the three isoflavone standards run with each sample batch. Peak areas were generally used to analyze the isoflavone concentrations quantitatively. The term ‘total isoflavone’ in this study refers to the sum of daidzein, genitein, and glycitein.

Determination of total phenolic compounds

Total phenolic compounds of soybean extracts were determined according to the previous method (Kozlowska et al., 1983; Zielinski and Kozlowska, 2000) with minor modifications. Two grams of ground soybean seed were mixed with 20 mL of 100% HPLC grade methanol in a 50 ㎖ screw-top flask, stirred for 24 h at 30℃, and filtered through a Whatman No. 42 filter paper. The filtrate was dried in a vacuum rotary evaporator (Eyela, Tokyo, Japan) below 30℃. Dried sample extracts were redissolved in 1 ㎖ of 100% methanol. A 0.25 ㎖ methanol extract of soybean was mixed with 4 ㎖ of distilled water and 0.25 ㎖ of Folin-Ciocalteu reagent, vortexed for 30 s, and left to stand for 5 min. The mixture was then added to 0.5 ㎖ of saturated sodium carbonate and allowed to stand for 30 min in the dark.

The absorbance was measured by Lambda 35 UV/vis spectrometer (Perkin-Elmer, Waltham, MA, USA) at 725 ㎚ against distilled water as blank. Each sample was analyzed in triplicate, and the results were expressed as gallic acid equivalents (milligrams of gallic acid equivalents per gram of sample) through the calibration curve of gallic acid.

Free radical scavenging activity

The free radical scavenging activities of soybean cultivars were determined using the DPPH scavenging method (Ozcelik et al., 2003). DPPH was dissolved in 100% methanol to a concentration of 0.5 mM. A total of 3.75 ㎖ of 0.5 mM DPPH was then mixed with 0.25 ㎖ of methanol extract as prepared above. The reaction mixture was maintained in the dark at room temperature for 30 min. The absorbance at 517 ㎚ was then recorded by Lambda 35 UV/vis spectrometer (Perkin-Elmer, Waltham, MA, USA). The free radical scavenging activity was calculated by the following equation:

Scavenging activity (%) = [Aa – (Ab – Ac)] / Aa × 100,

where Aa is the absorbance of the incubation solution of DPPH without the sample solution, Ab is the absorbance of the incubation mixture containing both sample solution and DPPH, and Ac is the absorbance of the blank solution without DPPH.

On-line detection of radical scavenging activity

The radical scavenging activity of the soybean seed was determined using the on-line ABTS+ assay according to the method of Stewart et al. (2005) with some modification. A 2 mM ABTS+ stock solution containing 3.5 mM potassium persulphate was prepared and incubated overnight in the dark at room temperature, to allow for the stabilization of the radicals. The ABTS+ reagent was prepared by diluting the stock 8-fold in methanol. Ten milligrams of each sample extract was dissolved in 1 ㎖ of methanol and filtered in a 0.2 ㎛ Acrodisc syringe filter (Pall, Lawrence, KS, USA), and 10 ㎕ of this solution was injected to an Agilent 1200 HPLC system (Agilent Technologies, Santa Clara, CA, USA) equipped with binary pumps, a diode array detection (DAD), a UV/VIS detector, and an additional reagent pump. The analytical column was a Hydrosphere-C18 column (5 ㎛ particle size, 150 ㎜ × 4.6 ㎜, YMC Co., Kyoto, Japan). The mobile phase consisted of acetonitrile (solvent A) and water (solvent B). A linear gradient (from 5 to 95% A for 60 min and 95% A for 10 min) was used. The chromatographic profile was recorded at 210 and 320 ㎚. The flow rate was 0.5 ㎖/min. The analyses were performed at 40℃. The HPLC eluent from the DAD arrived at a ‘T’ piece, where the ABTS+ was added. The ABTS+ flow rate was 0.25 ㎖/min delivered by an additional Agilent 1200 pump. After mixing through a 1-㎖ loop maintained at 40℃, the absorbance was measured at 734 ㎚ by a UV/VIS detector. Finally, the data were analyzed using ChemStation software (Agilent Technologies, Santa Clara, CA, USA).

Identification of compounds by HPLC-ESI/MS and NMR

The compounds were identified by a Varian 1200L LC-MS system (Palo Alto, CA, USA) with an electrospray source. The Varian MS workstation software (version 6.3) was used for data acquisition and processing. The HPLC conditions were the same as described above for the on-line detection of radical scavenging activity. The mass spectrometer conditions were as follows: positive ion mode, mass range = m/z 150-1,000, needle = 5,000 V, shield = 600 V, nebulizing gas pressure (N2) = 60 psi, drying gas (N2) flow rate = 20 psi, and drying temperature = 300℃.

Four major peaks were isolated by JASCO semi-preparative HPLC system (JASCO Co. Tokyo, Japan), which is equipped with PU-2086 Plus pumps, an MD-2015 PDA detector, and an MX2080-31 solvent mixing module. Separations were performed on a Hydrosphere-C18 column (5 ㎛ particle size, 250 ㎜ × 10 ㎜, YMC Co., Kyoto, Japan) with acetonitrile-water (5% acetonitrile at 0 min, 95% acetonitrile at 50 min) at a flow rate of 10 ㎖/min.

1H- and 13C-NMR data were acquired using a 500 MHz Varian NMR system (Varian, Palo Alto, CA, USA) equipped with a cold probe. Data were processed by Vnmrj software.

Tryptophan (1) ESI-MS m/z: 205 [M+H]+. 1H-NMR (500 MHz, DMSO-d6) δ: 2.92 (1H, dd, J=15.1, 9.8 Hz, H-10a), 3.29 (1H, dd, J=15.1, 3.8 Hz, H-10b), 3.39 (1H, dd, J=9.1, 3.8 Hz, H-11), 6.98 (1H, t, J=7.7 Hz, H-5), 7.06 (1H, t, J=7.7 Hz, H-6), 7.18 (1H, d, J=2.1 Hz, H-2), 7.34 (1H, d, J=7.8 Hz, H-7),7.37 (2H, brs, NH2) 7.55 (1H, d, J=7.8 Hz, H-4), and 10.9 (1H, brs, NH). 13C-NMR (125 MHz, DMSO-d6) δ: 27.7 (C-10), 55.3 (C-11), 109.7 (C-3), 111.7 (C-7), 118.7 (C-5), 118.8 (C-4), 121.4 (C-6), 124.4 (C-2), 127.7 (C-9), 136.8 (C-8), and 170.8 (C-12).

Epicatechin (2) ESI-MS m/z: 291 [M+H]+. 1H-NMR (500 MHz, DMSO-d6) δ: 2.46 (1H, dd, J = 16.4, 3.1 Hz, H-4a), 2.67 (1H, dd, J = 16.4, 4.5 Hz, H-4b), 3.99 (1H, ddd, J = 4.5, 3.1, 1.1 Hz, H-3). 4.68 (1H, brs, OH-3), 4.72 (1H, d, J = 1.1 Hz H-2), 5.71 (1H, d, J = 2.3 Hz, H-8), 5.89 (1H, d, J = 2.3 Hz, H-6), 6.64 (1H, dd, J = 8.2, 1.7 Hz, H-6′), 6.66 (1H, d, J = 8.1 Hz, H-5′), 6.88 (1H, d, J = 1.7 Hz, H-2′), and 8.80-9.12 (4H, brs, OH-5/7/3′/4′). 13C-NMR (125 MHz, DMSO-d6) δ: 28.7 (C-4), 65.3 (C-3), 78.5 (C-2), 94.5 (C-8), 95.5 (C-6), 98.9 (C-10), 115.2 (C-2′), 115.4 (C-5′), 118.4 (C-6′), 131.0 (C-1′), 144.9 (C-3′), 145.0 (C-4′), 156.2 (C-9), 156.7 (C-7), and 157.0 (C-9).

Daidzin (3) ESI-MS m/z: 417 [M+H]+. 1H-NMR (500 MHz, DMSO-d6) δ: 3.18 (1H, m, H-4″), 3.30 (2H, m, H-2″/3″), 3.45 (2H, m, H-5″/6″a), 3.71 (1H, m, H-6″b), 4.64 (1H, brs, OH-6″), 5.11 (1H, d, J = 7.4 Hz, H-1″), 5.13 (1H, brs, OH-4″), 5.19 (1H, brs, OH-3″), 5.48 (1H, brs, OH-2″), 6.82 (2H, m, H-3′/5′), 7.14 (1H, dd, J = 8.9, 2.3 Hz, H-6), 7.24 (1H, d, J = 2.3 Hz, H-8), 7.41 (2H, m, H-2′/6′), 8.05 (1H, d, J = 8.9 Hz, H-5), 8.40 (1H, s, J = 8.1 Hz, H-2), and 9.59 (1H, brs, OH-4′). 13C-NMR (125 MHz, DMSO-d6) δ: 61.1 (C-6″), 70.0 (C-4″), 73.6 (C-2″), 76.9 (C-3″), 77.7 (C-5″), 100.4 (C-1″), 103.8 (C-8), 115.4 (C-3′/5′), 116.0 (C-6), 118.9 (C-10), 122.7 (C-1′), 124.1 (C-3), 127.4 (C-5), 130.5 (C-2′/6′), 153.8 (C-2), 157.5 (C-4′), 157.7 (C-9), 161.8 (C-7), and 175.2 (C-4).

Genistin (4) ESI-MS m/z: 433 [M+H]+. 1H-NMR (500 MHz, DMSO-d6) δ: 3.16 (1H, m, H-4″), 3.29 (2H, m, H-2″/3″), 3.45 (2H, m, H-5″/6″a), 3.70 (1H, m, H-6″b), 4.63 (1H, m, OH-6″), 5.07 (1H, d, J = 7.5 Hz, H-1″), 5.12 (1H, d, J = 4.3 Hz, OH-4″), 5.21 (1H, brs, OH-3″), 5.45 (1H, d, J = 4.4 Hz, OH-2″), 6.47 (1H, d, J = 2.0 Hz, H-6), 6.72 (1H, d, J = 2.0 Hz, H-8), 6.83 (2H, m, H-3′/5′), 7.40 (2H, m, H-2′/6′), 8.43 (1H, s, J = 8.1 Hz, H-2), 9.68 (1H, brs, OH-4′), and 12.95 (1H, brs, OH-5). 13C-NMR (125 MHz, DMSO-d6) δ: 61.0 (C-6″), 70.0 (C-4″), 73.5 (C-2″), 76.9 (C-3″), 77.7 (C-5″), 94.9 (C-8), 100.0 (C-6), 100.3 (C-1″), 106.9 (C-10), 115.5 (C-3′/5′), 121.4 (C-3), 123.0 (C-1′), 130.6 (C-2′/6′), 155.2 (C-2), 157.7 (C-9), 158.0 (C-4′), 161.8 (C-5), 163.5 (C-7), and 180.8 (C-4).

Statistical analysis

The soybean was cultivated using a completely randomized design and replicated three times at two locations. The analysis of the isoflavones by HPLC was repeated four times with three extracts in each cultivar. Free radical scavenging activity and total phenolic compounds were measured in triplicates. The data were analyzed statistically by analysis of variance (ANOVA). The Duncan’s multiple range tests (probability level 0.05) were used to analyze the difference between the means of the samples using the SAS program (Table 1 and Table 2).

Table 2.zTotal = daidzein + genistein + glycitein. ymg of gallic acid equivalent per g of soybean. xData in the same column with different letters are significantly different at p < 0.05. w*, **Mean squares significant at the 0.05 and the 0.01 probability levels, respectively. vMean squares not significant at the 0.05 probability level.

 

Results and Discussion

Isoflavones

In soybeans, isoflavones are present both as aglycones and as glycosides (β-glycosides, acetyl glycosides, and malonyl glycosides), the latter being the predominant form (Lee et al., 2003; Lee et al., 2004; Kim et al., 2005). The biosynthetic pathway of isoflavones that leads to the accumulation of individual isoflavone compounds does not differ a great deal, although the total isoflavone content is influenced by the environment, cultivar, and their interaction (Sakthivelu et al., 2008). For this reason, we selected to measure isoflavone aglycone content following acid hydrolysis to investigate the impact of cultivar and planting region on the variation in the isoflavone content of soybean seeds.

The isoflavone content of the ten soybean cultivars grown at the two regions in Korea are shown in Fig. 1 and Table 2. The range of total isoflavone was 720 to 2,285 ㎍/g, which is similar to that reported in previous studies conducted in Korea (So et al., 2001; Lee et al., 2002). At GN (altitude 20 m), the total isoflavone content ranged from 720 to 2,205 ㎍/g and the average total isoflavone content was 1,361 ㎍/g. At PC (altitude 600 m), the total isoflavone content ranged from 841 to 2,285 ㎍/g and the average total isoflavone content was 1,444 ㎍/g. Therefore, the total average concentration of isoflavones in soybean cultivated at PC was significantly higher than at GN. Among the cultivars examined in this study, DP showed the highest total isoflavone content at both growing regions and TK exhibited the lowest concentrations at both regions. Together, there was a wide range of variation in the total isofalvone content of the cultivars. Most of the cultivars grown at the sub-highland PC site contained higher isoflavone content than those grown at the lowland GN site. PS and GJ contained a higher concentration of isoflavones when they were cultivated at GN. These results indicate that the cultivar × region interaction had a significant impact on the total isoflavone content (Table 2).

Fig. 1.Isoflavone content of the ten soybean cultivars grown at two different locations in Korea. Each bar indicates the standard deviation values.

Planting location and genotype may explain the variation in total isoflavone content. Hoeck et al. (2000) and Riedl et al. (2007) previously confirmed the influence of planting location and cultivar on the total isoflavone content at four sites in Ohio and at eight sites in Iowa, respectively. Another study reported that low growing temperature during the first few weeks after seeding promote isoflavone biosynthesis in soybean (Tsukamoto et al., 1995). In common with previous studies, we found that isoflavone concentrations are influenced by various environmental factors, such as cultivation site and temperature, as well as by genetically controlled factors, such as cultivar, seed weight, and seed color (Hoeck et al., 2000; Lee et al., 2003).

Total phenolic compounds

We found that the total phenolic content ranged from 2.2 to 6.9 ㎍/g (Fig. 2). The total phenolic content of the cultivars grown at GN was 2.5 to 6.9 ㎍/g, with an average of 4.2 ㎍/g. At PC, the total phenolic content ranged from 2.5 to 4.4 ㎍/g, with an average of 3.3 ㎍/g (Table 2).

Fig. 2.Total phenolic content of the ten soybean cultivars grown at two different locations in Korea. Each bar indicates the standard error values.

All of the cultivars grown at GN had greater total phenolic content than those grown at PC. Among the cultivars studied, JN had the highest concentration of total phenolic content at both growing locations. In particular, JN grown at GN had much higher total phenolic content (6.9 ㎎/g) than any of the other samples. TK exhibited the lowest total phenolic content of all the cultivars at both sites. The other eight cultivars exhibited little variation in total phenolic content, and there was no significant difference between these cultivars in total phenolic contents (Table 2). The trend in total phenolic content contrasted with our findings for isoflavone concentrations, which were higher in cultivars grown at PC where was also a significant difference between the cultivars in total isoflavone concentrations.

As isoflavones are phenolics and likely a major contributor to the total phenolic content, it is natural to assume that the total phenolic content is closely associated with the total isoflavone content. Some researchers have reported that the total phenolic content is positively correlated with the total isoflavone concentration (Riedl et al., 2007; Sakthivelu et al., 2008; Devi et al., 2009). However, we did not find any such correlation between total phenolics and isoflavones. It is suggested more phenolic compounds except isoflavone are included in soybeans cultivated in GN. The isoflavone content of soybean cultivars used in this study was higher at PC, whereas the total phenolic content was greater at GN. Among the cultivars studied, the highest isoflavone content was observed in DP; while the highest total phenolic content was detected in JN.

Previous studies have reported that soybean contains several phenolic acids, such as syringic, ferulic, sinapic, coumaric, hydroxybenzoic, caffeic, and chlorogenic acids (Kozlowska et al., 1983) and that total phenolic compounds are highly positively correlated with hydroxybenzoic acid, such as gentisic acid and salicylic acid (Kim et al., 2004). Lee et al. (2008) reported that soybean cultivars grown in Korea contain 28 phenolic compounds including eleven hydroxybenzoic acids, seven hydroxycinnamic acids, eight flavonoids, and two others. Considering that several phenolic compounds are present in soybean, it is feasible that the total phenolic content is not always correlated with the total isoflavone content. Further studies are needed to investigate the relationship between total phenolics and isoflavones and to identify all of the phenolic compounds present in soybean.

Antioxidant activity

The antioxidant activities of the ten soybean cultivars grown at the two locations are shown in Fig. 3 and Table 2 as a percentage of the radical scavenging activity of DPPH. There were significant differences in the free radical scavenging activities of the cultivars (P < 0.05). JN cultivated at GN exhibited the highest free radical scavenging activity (67.8%) whereas PS cultivated at PC had the lowest (30.6%). The average free radical scavenging activities of the ten soybean cultivars grown at GN and at PC were 45.5 and 39.3%, respectively. There was a significant difference in DPPH radical scavenging activity between the two sites. Our findings are similar to those of Lee et al. (2002), who concluded that the free radical scavenging activities of soybean extracts are influenced by genetic and environmental variations.

Fig. 3.Free radical scavenging activity of the ten soybean cultivars grown at two different locations in Korea. Each bar indicates the standard error values.

Overall, the trend in DPPH radical scavenging activity was similar to that of total phenolic content rather than that of total isoflavone content. Cultivars with the lowest total phenolic content, such as TK and PS exhibited only about half of the antioxidant activity of the phenolic-rich cultivar, JN. Despite having low total isoflavone content, JN, CJ, and IP showed high levels of DPPH radical scavenging activity.

Both isoflavone aglycones and glycosides were present in soybean, with glycosides as the predominant isoflavone form (Lee et al., 2003; Lee et al., 2004; Kim et al., 2005). Previous research suggested that glucose linkage to aglycone reduced the antioxidant activities of isoflavones approximately 50 to 100 times (Naim et al., 1976). Moreover, Cos et al. (2003) reported that isoflavones were less potent in a DPPH assay compared with certain phenolic acids present in soybean. Lee et al. (2005) suggested that many isoflavones have low scavenging potency for DPPH free radicals, with scavenging effects only half that of α-tocopherol and one-third that of epicatechin. Therefore, it is possible that phenolic acids or some other compounds present in soybean seeds were responsible for much of the DPPH radical scavenging activity detected in this study.

On-line HPLC-ABTS+ radical scavenging activity and the identification of compounds by HPLC-ESI/MS and NMR

Many methods, such as ABTS and DPPH assay have been used to evaluate the free radical scavenging capacity of compounds. However, these methods cannot determine the active components in a mixture. Recently, a technique has been developed that measures the radical scavenging activity of individual compounds on-line when they elute from the HPLC column (Koleva et al., 2001; Lee et al., 2007). This technique allow for rapid and selective detection of radical scavenging compounds in the presence of many other inactive constituents with minimal sample preparation. We used HPLC coupled with the on-line ABTS+-based assay system and HPLC-ESI/MS to rapidly identify radical scavenging compounds in soybean extract.

An example of the antioxidant analysis of soybean extract is shown in Fig. 4. The upper part (positive) of the figure shows the UV chromatogram as measured at 210 (dotted) and 320 ㎚ (solid line). The down part (negative) of the figure shows the corresponding ABTS+ radical scavenging activities of the same compounds. The chromatographic analyses of the soybean extract at 210 and 320 ㎚ revealed the existence of four major compounds, which were identified as tryptophan (1), epicatechin (2), daidzin (3), and genistin (4) by LC-MS and NMR (Fig. 5). Among the four compounds, epicatechin exhibited the strongest antioxidant activity.

Fig. 4.On-line HPLC-ABTS+ analysis of soybean extract. Ten milliliters were analyzed by gradient reverse-phase HPLC with a DAD at 210 ㎚ (positive trace, dotted line) and 320 ㎚ (positive trace, solid line) prior to reaction with the ABTS+ radical and the analysis of antioxidant potential at 734 ㎚ (negative trace). (tryptophan (1), epicatechin (2), daidzin (3), and genistin (4)).

Fig. 5.Structures of the four major compounds present in the soybean extract.

The UV spectrum of peak 2 showed maximum absorbance at 225 and 277 ㎚. The mass analysis of peak 2 showed a [M+H]+ ion at m/z 291 for ESI-MS in the positive ion mode. Based on the UV and MS spectral data and previously reported literature (Zeeb et al., 2000; Qian et al., 2008; Vasco et al., 2009), peak 2 was identified as epicatechin. The NMR data showed that coupling between H-2 and H-3 was negligible (J = 1.1 Hz), thereby confirming that peak 2 was epicatechin. If there is a trans relationship between C-2 and C-3, large coupling (J = 8.0 Hz) should be observed (Hsue-Fen et al., 1993).

Catechin and epicatechin are flavonoids that are widespread in many edible plants such as green tea. Numerous studies have addressed the antioxidative activities of these flavonoids in scavenging free radicals, chelating metal ions, preventing the activation of redox-sentitive transcription factors, and inhibiting prooxidant enzymes (Medina et al., 2007; Stangl et al., 2006). In particular, epicatechin was reported to exhibit antioxidant activity three times higher than that of many isoflavones (Lee et al., 2005). Lee et al. (2008) reported that catechin is present in soybean seed and Cho et al. (2009) reported that both catechin and epicatechin are present in soybean-fermented food. However, the antioxidant activity of epicatechin in soybean has not been reported previously. The antioxidant detection system used in this study clearly indicates for the first time that epicatechin makes a major contribution to the total antioxidant activity of soybean seeds.

In this study, we showed total isoflavones, total phenolics, and antioxidant activity of soybeans differed by locations (altitude) or cultivars. The antioxidant activity could be largely linked to total phenolics rather than isoflavones. Correlations among phenolic content, location and genotype found in this study can help distinguish variables associated with antioxidant activity of soybeans. However, more environmental variations such as temperature, precipitation, humidity and sunshine need to be investigated.

In addition, the on-line HPLC-ABTS+ method was applied successfully to evaluate the ABTS-radical scavenging activity of each peak concurrently. We demonstrated that epicatechin is the most potent antioxidant constituent comprised in those soybean seeds.