|Home | About | Journals | Submit | Contact Us | Français|
Gastrointestinal (GI) mimic and organic solvent extracts of whole soybean powder (WSP), soy protein concentrate (SPC), and soy protein isolate (SPI) as well as soy isoflavone concentrate (SIC) were analyzed for total phenols; quinone reductase (QR) induction in hepa1c1c7 cells; antioxidant scavenging of DPPH, HOCl, ONOO−, and O2−.; and total antioxidant capacity via FRAP and ORAC assays. GI extracts of all the soy products had higher concentrations of total phenols than from acidified methanol (MeOH) but lower antioxidant potency. The MeOH extract of SPC was most potent in quenching HOCl and ONOO− and increasing FRAP and ORAC, but did not induce QR. Despite weak antioxidant activity, hexane (HX) extracts induced QR more than GI and MeOH extracts with WSP > SPC > SPI > IC. Soy extracts were ineffective scavengers of DPPH and O2−.. Thus, extraction methods markedly affect the antioxidant profile and QR induction capacity of soy products.
Soybean and its derived products are used extensively for food formulation, including meat analogues, infant formula, and soymilk, as well as dietary supplements. Soy is considered a desirable ingredient for health promotion, particularly because of its inverse association in observational studies with the risk of cardiovascular disease, some forms of cancer, menopausal symptoms, and osteoporosis (Omoni & Aluko, 2005). These health benefits have been ascribed substantially to the bioactivity of soy polyphenols, particularly the isoflavones, including their antioxidant, anti-proliferative, and hypocholesterolemic effects as well as binding to estrogen receptors (Isanga & Zhang, 2008). Isoflavones have been shown in a variety of in vitro and animal studies to be potent antioxidant and chemopreventive flavonoids (Messina & Flickinger, 2002; Rufer & Kulling, 2006; Omoni & Aluko, 2005). The efficacy of isoflavones is dependent on their relative enrichment in soy products, which can vary as much as 50-fold due to processing methods (Wang & Murphy, 1994).
Rapid, in vitro screening methods have been employed to test the potential bioactivity of plant foods; e.g., an array of free radical scavenging assays has been used to identify and rank foods by their antioxidant capacity (Paganga, Miller, & Rice-Evans, 1999). Induction of phase II detoxification enzymes, such as quinine reductase (QR), has been employed to identify foods with chemoprotective potential (Zhang, Talalay, Cho, & Posner, 1992). Of course, the relevance of results from these in vitro tests to bioactivity in vivo is always limited due to their independence from nutrient factors such as bioaccessibility, bioavailability, and metabolism. However, a greater degree of predictability from in vitro antioxidant assays may be possible by more directly mimicking physiology and testing specific reactions found in vivo.
In screening plant foods for antioxidant activity, organic solvents are commonly employed to maximize the extraction of phytochemicals for testing against various chemical probes. For example, the in vitro antioxidant activity of soy has been assessed against the synthetic diphenyl-1-picryl-hydrazyl (DPPH) radical following extraction with acetone (Xu & Chang, 2007; Xu, Yuan & Chang, 2007) and acetonitrile (Lee, Renita, Fioritto, St. Martin, Schwartz, & Vodovotz, 2004). However, this approach does not reflect the “extraction” or absorption processes in the gastrointestinal tract and the free radicals generated in the cell milieu.
Our hypothesis was that a gastrointestinal extraction method mimicking the “solvent” of the gastrointestinal lumen and organic solvent extracts of soy products have different in vitro antioxidant capacity and bioactivity. Therefore, the aim of this study was to determine the extent to which a gastrointestinal-simulated extraction (GI) differs from extraction with acidified methanol (MeOH) and hexane (HX) and affects the antioxidant capacity and QR induction in vitro of whole soy powder (WSP), soy protein concentrate (SPC), and soy protein isolate (SPI). We employed a soy isoflavone concentrate (SIC) as a reference standard to assess the potential contribution of isoflavones to bioactivity. The radical scavenging capacity of these soy extracts was measured in vitro against the physiologically relevant reactive species peroxynitrite (ONOO−), superoxide anion (O2−.), and hypochlorite (HOCl), as well as against the commonly used DPPH radical. In addition, the “total antioxidant capacity” of the extracts in vitro was determined by the ferric reducing antioxidant power (FRAP) and oxygen radical absorbance capacity (ORAC) assays.
Dubelco’s modified eagle medium (DMEM) was obtained from American Type Culture Collection (Manassas, VA) and fetal bovine serum from Hyclone (Logan, UT). ONOO− in 0.3 N NaOH was purchased from Cayman (Ann Arbor, MI); 2,2′-azobis (2-amidinopropane) dihydrochloride (AAPH) from Wako Chemicals USA (Richmond, VA); NaOH, methanol, and HCl from Fisher Scientific (Pittsburgh, PA). All other chemicals were acquired from Sigma Chemical (St. Louis, MO) or Sigma-Aldrich (St. Louis, MO). SPC (ARCON® S; containing 72% protein), SPI (PRO-FAM 955; containing 90% protein), WSP (NutriSoy®; containing 43% protein and 19.3% fat), and SIC (NOVASOY 400; containing 43.5% isoflavones) were generously provided by ADM (Decatur, IL).
An accelerated solvent extraction system (ASE200, Dionex Corp., Sunnyvale, CA) was used to obtain MeOH and HX extracts of SPI, SPC and WSP, according to the method of Chen and Blumberg (2008). Briefly, powdered soy samples (3 g) were mixed with an equivalent mass of diatomaceous earth and loaded into 11 mL extraction cells topped off with sand. For both MeOH and HX extractions, the flush volume was set to 50%, purge at 180 s, pressure at 1500 psi, static cycle at 5 min, and temperature at 100°C. MeOH extraction was performed using a sequential extraction with 90, 60, and 30% methanol solutions containing 5% acetic acid. HX extraction was run for 4 cycles with 100% HX. After the exact volume of extracts was measured and recorded, they were centrifuged for 5 min at 4000 × g, and divided into aliquots for subsequent assays.
GI extracts of soy products were performed to simulate the digestion process, according to the method of Chen and Blumberg (2008). Briefly, powdered soy sample (1 g) was mixed with a saline solution and digested with pepsin (pH = 2) in a shaking water bath at 37°C for 1 h. The solution was then incubated with a pancreatin-bile solution at pH 6.9 for 2 h in a shaking water bath at 37°C. Pepsin and pancreatin may interfere with antioxidant activity assays, so aliquots of the GI extracts were mixed with equal volumes of 100% methanol to precipitate protein, centrifuged, and the supernatant divided into aliquots for subsequent assays.
Aliquots of the 3 extracts were dried under N2 gas at room temperature and stored at −20°C. Prior to the assays, MeOH extracts were reconstituted in 60% methanol and 5% acetic acid, HX extracts in acetic acid, and GI extracts in water, respectively. SIC was not extracted, but rather dissolved directly in methanol.
Total phenols of reconstituted extracts were determined according to the modified methods of Singleton, Orthofer, and Lamuela-Ravent (1999) with results expressed as μM gallic acid equivalents (GAE) or mg soy product/100 μmol GAE. For subsequent antioxidant and QR assays, doses of SPI, SPC, WSP, and SIC at 100, 10, and 0.1 μM GAE were used to reflect the range of concentrations potentially present in the gastrointestinal tract, plasma, and cells, respectively, following ingestion.
DPPH scavenging activity was performed according to Brand-Williams, Cuvelier, and Berset (1995). Briefly, DPPH in ethanol was mixed with an equal volume of different concentrations of the soy extracts and the absorbance at 520 nm was measured after 30 min incubation at room temperature in the dark. Intra- and inter-day assay coefficients of variation (CV) were 1.4 and 7.6%, respectively.
Scavenging activity against ONOO− was measured by monitoring the increase in fluorescence from the oxidation of dihydrorhodamine 123 (DHR123) according to a slightly modified method of Choi, Choi, Han, Bae, and Chung (2002). The concentration of ONOO− stock solution was determined by a spectrophotometry after alkalization using a cold 0.3 mol/L NaOH solution at a ratio of 1:40, and aliquots were stored at −80°C. Immediately before use, ONOO− was diluted to a final concentration of 100 μM. Fluorescence at 485 nm excitation and 530 nm emission generated from DHR123 oxidation 5 min after the addition of ONOO− was recorded using a FLUOstar Optima multifunctional plate reader (BMG Labtech Inc. Durham, NC). Intra- and inter-day assay CV were 4.7 and 3.6%, respectively.
Scavenging activity against HOCl was assessed via the oxidation of ferrocyanide [Fe(II)CN)6] in a phosphate buffer as a reference reaction to investigate the stoichiometry of the reaction according to modification of the methods described by Zhu, Carr, and Frei (2002) and Prutz (1996). The concentration of HOCl stock solution obtained from Sigma was determined according to Hussain, Trudell, and Repta (1970). Briefly, the soy extracts were incubated with HOCl for 5 min at room temperature before the addition of Fe(II)(CN)6 and then absorbance was monitored at 420 nm using a Shimadzu UV1601 spectrophotometer (Japan). Intra- and inter-day assay CV were 0.9 and 2.9%, respectively.
Scavenging activity against O2−. was measured in a xanthine/xanthine-oxidase system with spectrophotometric determination of the reduction product of nitroblue tetrazolium (NBT) according to a slight modification of the method described by Chun, Kim, and Lee (2003). Briefly, following 10 min of incubation of the soy extracts at room temperature with a reaction mixture of 50 μM NBT, 50 μM xanthine, and 0.05 U/mL xanthine oxidase (final concentrations), the change in absorbance of NBT was measured at 560 nm using a Shimadzu UV1601 spectrophotometer. Intra- and inter-day assay CV were 1.9 and 7.7%. Inhibition of xanthine oxidase activity by the extracts was monitored by the spectrophotometric determination of uric acid production.
Results of radical scavenging activity are expressed as a percentage of the control (no soy extract present) and the IC50 (concentration of the soy extract required to decrease absorbance by 50%) in μM GAE, calculated using a spline function.
The “total antioxidant capacity” of the extracts was assessed using the ORAC and FRAP assays. The ORAC assay was conducted according to Ou, Hampsch-Woodill, and Prior (2001). The ORAC assay employs the area under the curve of the magnitude and time to the oxidation of fluorescein due to peroxyl radicals generated by the addition of AAPH. The assay was carried out on a FLUOstar OPTIMA plate reader utilizing fluorescence filters with 485 nm excitation and 520 nm emission. All data are expressed as μmol Trolox Equivalents (TE)/μmol GAE. Intra- and inter-day assay CV were 3.0 and 7.3%, respectively.
The FRAP assay determines the capacity of antioxidants as reductants in a redox-linked colorimetric reaction of the reduction of Fe3+-2,4,6-tri-pyridyl-S-triazine to a blue-colored Fe2+ complex at low pH that is measured spectrophotometrically at 593 nm (Benzie & Strain, 1996). The extracts were incubated at room temperature with the FRAP reagent and the absorbance recorded after 1 h (Chen & Blumberg, 2008). The reducing power is expressed as μmol TE/μmol GAE. Intra- and inter-day assay CV were 0.7 and 4.2%, respectively.
The modulation of QR activity in murine hepatoma Hepa1c1c7 cells has been widely employed as a tool to examine the potential chemopreventive activity of phytochemicals (Kang & Pezzuto, 2004). Hepa1c1c7 cells were cultured until confluent in DMEM supplemented with 10% heat inactivated, charcoal-treated fetal bovine serum, 1% penicillin/streptomycin, and 1% L-glutamine in an incubator with 5% CO2 at 37°C. After confluence, cells were plated at a concentration of 2×104 cells/well in 96-well clear plates and allowed to settle for 24 h. After medium was aspirated, cells were treated with the soy extracts in the medium for 48 h. QR activity was measured by an NADPH-generating system, coupling the oxidation of menadione to the reduction of the dye 3-(4,5-dimethylthiazol-2-yl)-2,5 diphenyl tetrazolium bromide according to Kang and Pezzuto (2004). The resulting blue-brown color was measured at 570 nm using a FLUOstar Optima plate reader. The protein content of cells in each well was determined with a BCA protein kit (Pierce, Rockford, IL). Following adjustment of protein content, QR activity was expressed as nmol/mg protein/min. β-Napthoflavone at concentration of 1 μM was employed as a positive control and increased QR activity 2.4 ± 0.5-fold of the negative control (absent soy extract).
All results are expressed as mean ± SD. Statistical comparisons between extraction method and soy products were performed by 2-factor ANOVA. Differences in antioxidant and QR activity among soy products were analyzed by 1-factor ANOVA, followed by post hoc analysis using Fisher’s protected Least Significant Difference (LSD) test. Pearson’s correlation tests were performed to reveal possible associations between the antioxidant activity assays, using the JMP IN 4 statistical software package (SAS Institute Inc., Cary, NC). Differences with P <0.05 were considered significant.
From 0.1 to 100 μM GAE, SPI, SPC, and WSP extracts and SIC had weak to no antioxidant activity in the DPPH and O2−. scavenging assays (data not shown). The antioxidant activity of a methanolic soy extract against DPPH has been reported to be 7 μmol TE/g (Xu & Chang, 2007). Gerhäuser et al. (2003) observed that IC50 of DDPH and O2−. scavenging activities for the isoflavone genistein was >250 and >100 μM, respectively. Therefore, it is likely in the present study that the soy extract doses approximating bioavailability were too low to scavenge DPPH and O2−..
Extraction methods had a significant effect on total phenols content of the soy products using 2-factor ANOVA (P = 0.02), with the rank of total phenols being GI > MeOH > HX. A consistent relationship between the GI, MeOH, and HX extraction methods and the other assays utilized here was not apparent.
In contrast to the general understanding that extraction by organic solvents affords greater recovery of total phenols than aqueous solvents, the GI extract liberated more total phenols than MeOH and HX in all soy products (Table 1). Consequently, the GI extracts yielded products with the largest antioxidant quantity (i.e., antioxidant potency expressed on a mass basis) in the ONOO−, ORAC, and FRAP assays than MeOH and HX extracts for all soy products tested (Table 2). Since GI extraction employed protease digestion, this finding is in agreement with previous studies in which hydrolysis of soy protein improved antioxidant efficacy due to liberation of bound phenols, release from chelating agents such as phytic acid, and production of antioxidant peptide sequences (Yee, Shipe, & Kinsella, 1980; Chen, Muramoto, & Yamauchi, 1995; Saito, et al. 2003).
Despite a higher antioxidant quantity, the antioxidant quality (i.e., potency expressed on the basis of total phenols) of GI extracts was moderate in the ONOO−, ORAC, and FRAP assays relative to HX and MeOH. In contrast, GI extracts of SPI and WSP possess high antioxidant quality toward HOCl, with a 24% greater potency than the MeOH extracts. These results demonstrate that antioxidant profiles of soy products are dependent upon the solvent used to extract them, an effect that is most likely due to the different phytochemical profiles of the respective extracts. The content of total phenols and flavonoids in soybeans and their effects on the FRAP, ORAC, and DPPH assays have been previously demonstrated to change with extraction solvent choice (Xu & Chang, 2007). Further, using extracts of almond skins, we found the flavonoid profiles of MeOH and GI extracts to be markedly different (Chen & Blumberg, 2008).
HX extracts were found to possess an antioxidant quality in the order of SPC > SPI > WSP in the FRAP assay. However, HX extracts of SPI, SPC and WSP yielded lower concentrations of total phenols than GI or MeOH extracts, possessed weaker antioxidant quantity in the FRAP and ORAC assays, and did not exhibit antioxidant activity against ONOO− and HOCl.
The difference in assay results between the GI extracts compared to the organic solvent extracts highlights the substantial impact of this process on ranking the capacity of soy products with regard to their antioxidant quality and quantity. A limitation to this approach is that in vivo metabolism and biotransformation, which may have profound effects on the in vitro antioxidant activity of soy constituents (Rufer & Kulling, 2006), are not accounted for. Therefore, careful consideration is required in the design and interpretation of such studies to better allow an extrapolation of the results to potential in vivo bioactivity.
The concentration of total phenols from the soy products provided from MeOH extracts ranked as WSP > SPI > SPC, reflecting an inverse association with the relative protein purity in these soy products. The production of SPI and SPC can include alcohol washing, protein denaturation, and/or precipitation plus concentration, processes that can enhance the loss of phenolic compounds. Consistent with our results, previous studies have reported decreased extractable phenols and isoflavones in SPI and SPC extracts (Wang & Murphy, 1994; Pinto, Lajolo, & Genovese, 2005).
MeOH extracts of SPC exhibited the greatest antioxidant quality against HOCl and ONOO− and possessed the highest FRAC and ORAC values. Similarly, the HX and GI extracts of SPC had greater antioxidant quality than SPI and WSP in the FRAP and ORAC assays. In contrast, the total phenols in SPC following MeOH extraction were 9-fold less than from WSP and SPI. The total phenols concentration of GI extracts of SPC, SPI and WSP were not markedly different, varying ≤31%. Interestingly, while the antioxidant quantity of SPC was less than SPI and WSP, likely due to its manufacturing process, this same process appears to have enriched its antioxidant quality.
Although the processing did not exert a consistent impact on antioxidant activity measured by an array of assays (P ≥0.05), the antioxidant quality of the MeOH extract of SPC was consistently greatest among the HOCl, ONOO−, FRAP, and ORAC assays when compared to the other soy products, including SIC. Thus, new efforts to isolate and identify the constituents responsible for the enhanced antioxidant quality of MeOH extracts of SPC are warranted.
Soybean extracts and individual isoflavones have been reported to inhibit DPPH, 2,2′-azinobis(3- ethylbenzothiazoline-6-sulphonic acid) diammonium salt (ATBS) radical cation, and oxygen centred radicals as well as the ex vivo oxidation of low density lipoprotein cholesterol (Xu & Chang, 2007; Ungar, Osundahunsi & Shimoni, 2003; Rufer & Kulling, 2006; Lai & Yen, 2002). We found the antioxidant quality of purified soy isoflavones was lower than that of the soy extracts in the HOCl assay, but more effective than most of the extracts in the ONOO− and ORAC assays (except MeOH extracts of SPC). While the contribution of isoflavones to the antioxidant activity of the soy extracts is greater in the ONOO− and ORAC assays, compared to the FRAP and HOCl assays, the importance of this observation to predicting in vivo bioactivity remains to be explored.
Determining the induction of QR activity provides an additional assessment of antioxidant function, as QR expression is induced via the redox-sensitive Antioxidant Response Element (ARE) in the genome (Talalay, De Long, & Prochaska, 1988). QR also possesses a more direct antioxidant function by participating in the reduction of the quinone form of oxidized α-tocopherol (Siegel, Bolton, Burr, Liebler, & Ross, 1997). However, neither the soy extracts nor SIC induced QR in Hepa1c1c7 cells in the tested range of 0.1 to 1 μM GAE. At 100 μM GAE, cell viability was reduced ≤50% for all the extracts. The MeOH extract of WSP at 10 μM GAE induced QR activity by 31%, whereas the same extract of SPI and SPC had no effect in this assay (Table 3). Thus, the relatively higher antioxidant quality of the MeOH extract of SPC was not correlated with the induction of QR activity. HX extracts of the soy products induced QR activity from 42 to 66% at 10 μM GAE. Since HX extracts possess the lowest antioxidant quality, constituents other than the soy polyphenols appear to be contributing partly to QR induction.
The isoflavone genistein, and, to a lesser extent, daidzein and their glycosides have been found to significantly induce QR activity in Hepa1c1c7 and HepG2 cells (Yannai, Day, Williamson, & Rhodes, 1998; Chun, Kim, & Lee, 2005). We found SIC, containing 43.5% isoflavones, induced QR activity by 1.3-fold at 10 μM GAE. However, SIC was 14–44% less effective at inducing QR than the HX extracts of SPI, SPC and WSP and the MeOH extract of WSP. Previous bioassay-guided fractionation of soy flour also found phenolic esters and other uncharacterized constituents were more potent QR inducers than isoflavones (Bolling & Parkin, 2008; Bolling & Parkin, 2009). Therefore, the capacity of the soy extracts to induce QR may be mediated via a synergistic interaction between their phenolic constituents or through other soy compounds in the extract that are more potent than isoflavones.
Several significant correlations between the antioxidant assays are evident when their results are expressed as antioxidant quantity (Table 4). The content of total phenols correlates positively with the results of the FRAP, ORAC, and HOCl assays (P ≤0.05). Notably, no correlations were found between the content of total phenols and antioxidant quality, suggesting that standardizing these assays based on their content of total phenols may not provide a worthwhile approach for such comparisons. This apparent discordance may reflect contributions to antioxidant capacity from compounds other than isoflavones and related polyphenols in soy extracts and/or variations in efficacy and potency of the soy polyphenols individually or in combination.
Extraction methods markedly affect the antioxidant profile and QR induction capacity of soy products. Measures of antioxidant quantity, with potency expressed on a mass basis, and antioxidant quality, with potency expressed per concentration of total phenols, appear to be useful measures for creating a more comprehensive profile of the antioxidant capacity of soy extracts in vitro. However, among the assays included in our screening panel for antioxidant capacity, no consistent effect of processing was apparent. This approach did reveal that among the SPC, SPI and WSP extracts and SIC, the isoflavones contributed most importantly to the quenching of ONOO− and the increase in total antioxidant capacity measured by the ORAC assay, but not to the induction of QR activity. Results from the other assays in our panel of tests suggest that soy constituents other than the isoflavones may contribute to the antioxidant actions of soy extracts in vitro. For example, MeOH extracts of SPC provided the highest antioxidant quality but had no effect on the induction of QR, while the HX extract, with its low concentration of total phenols, was the most potent inducer of QR activity. These results suggest that a broader range of extracts and a more comprehensive set of in vitro tests than are typically conducted on plant food extracts may be necessary to provide better prediction of potential in vivo antioxidant actions. As GI extracts provided higher concentrations of total phenols than those from organic extracts, the use of extraction methods that more closely resemble the conditions found in vivo should prove an important part of these evaluations.
Bradley Bolling was supported by NIH IRACDA Training Grant K12 GM074869. We thank Jennifer O’Leary, Kendra Hamel, and Desiree Kelley for their excellent technical assistance.
1Supported by Interagency Agreement No. 59-1950-7-735 between the National Institutes of Health (NIH)/National Center on Complementary and Alternative Medicine and U.S. Department of Agriculture (USDA)/Agricultural Research Service and USDA ARS under Cooperative Agreement No. 58-1950-7-707. The contents of this publication do not necessarily reflect the views or policies of the NIH or USDA nor does mention of trade names, commercial products or organizations imply endorsement by the U.S. government.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.