Analysis of Isoflavones in Soy Foods

Transcription

1 Analysis of Isoflavones in Soy Foods UNIT I1.6 Isoflavones are characteristically found in the family of the Leguminosae, serving such biofunctions as phytoalexins and regulating growth and development of plants (Ingham, 1982). Increasing evidence suggests that consumption of soybean products may significantly impact health (Setchell, 1998, 2001; Birt et al., 2001). The biological activity has been associated, in part, with the presence of isoflavones in soy (Fournier et al., 1998; Setchell and Cassidy, 1999). Analysis of these bioactive compounds in soybean products is an essential part of any research involving soy isoflavones. This unit attempts to provide a reliable method with the most commonly used analytical techniques for this purpose. Soybeans are the most prominent source of isoflavones in food plants, ranging in concentration from 0.1 to 3.0 mg/g dry weight (Coward et al., 1993). The general structure of isoflavones is shown in Figure I1.6.1, and the common names, chemical names, and chemical structures of the twelve isoflavones in soy are shown in Figure I Because isoflavones are present at small concentrations in soybeans and soy-based food products (see matrices in Table I1.6.1), their efficient extraction prior to analysis is very important. In this unit, extraction of the twelve soy isoflavones is described (see Basic Protocol 1), as is sample preparation before extraction (see Support Protocol 1). This process consists of removing water from the sample and sample grinding or homogenization. When authentic isoflavone standards are available, the method of choice for analyzing isoflavone extracts from soy foods is generally high-performance liquid chromatography (HPLC) paired with a reversed-phase C18 column and UV spectrophotometer. A mixture of isoflavones in soy food extracts is separated based on the polarity and/or solubility of isoflavones in the column between the stationary (packing material of the column) and mobile (solvent) phases used. All isoflavones exhibit an intense absorption in the UV region of the spectrum, between 240 and 280 nm (Harborne, 1967). Based on the Beer-Lambert law, the concentration of a pure isoflavone compound or an individual isoflavone in a mixture can be calculated using the absorbance measurements obtained and known extinction coefficients (llis, 1962). This unit describes techniques for generating calibration curves with reference standards (see Basic Protocol 2), separating and measuring isoflavones using HPLC-UV spectrophotometry (see Basic Protocol 3), preparation of samples for HPLC (see Support Protocol 2), and converting glycosidic isoflavones to their aglycones using enzymatic hydrolysis (see Support Protocol 3). NTE: All reagents used in this unit, including water, should be HPLC grade or equivalent. NTE: Isoflavones should be protected from light, oxygen, and elevated temperature during and after extraction (see Critical Parameters and Troubleshooting). H() (CH 3 )H A C (glyc) benzoyl 5 6 B cinnamoyl Figure I1.6.1 General structure and ring numbering system applied to known soy isoflavones. Contributed by Yu Chu Zhang and Steven J. Schwartz Current Protocols in Food Analytical Chemistry (2003) I1.6.1-I Copyright 2003 by John Wiley & Sons, Inc. I1.6.1

2 BASIC PRTCL 1 SLVENT EXTRACTIN F ISFLAVNES IN THEIR NATURAL FRMS FRM SY FDS This method focuses on the extraction of soy food samples at room temperature using acidified solvent mixtures. The method is designed to release isoflavones in their natural forms from food matrices based on their polarity and solubility in solvents. Insoluble proteins, carbohydrates, and lipids present in the food are removed from the isoflavone extract using organic solvents, acid, and centrifugation. Common name Chemical name Chemical structure Daidzein 7,4 -Dihydroxyisoflavone Daidzin Acetyldaidzin 7,4 -Dihydroxyisoflavone 7- glucoside or daidzein 7--glucoside or daidzein 7--β-D- glucopyranoside 6 --Acetyldaidzin H H Malonyldaidzin 6 --Malonyldaidzin H C H Genistein Genistin Acetylgenistin 5,7,4 -Trihydroxyisoflavone or 5,7-dihydroxy-3-(4- hydroxyphenyl)-4h-1- benzopyran-4-one 5,7,4 -Trihydroxyisoflavone 7-glucoside or genistein 7--glucoside or genistein-7--β-d- glucopyranoside 6 --Acetylgenistin H Malonylgenistin 6 --Malonylgenistein H C H Analysis of Isoflavones in Soy Foods I1.6.2 Figure I1.6.2 (above and at right) Common name, chemical name, and chemical structure of soy isoflavones. Current Protocols in Food Analytical Chemistry

3 Materials Prepared food sample (see Support Protocol 1) Acetonitrile 0.1 M HCl 50-ml centrifuge tube with cap Ultrasonic bath Wrist shaker 15-ml centrifuge tube Analytical pipets 1. In a 50-ml centrifuge tube with cap, carefully measure the prepared food sample and mix it with solvent (i.e., acetonitrile, HCl, and water) as detailed in Table I Break up cellular material using an ultrasonic bath for 10 min. 2. Vortex the mixture for 1 min to homogenize the suspension. 3. Extract on a wrist shaker 2 hr at room temperature ( 25 C). 4. Centrifuge the mixture 30 min at 430 g, room temperature. 5. Transfer an aliquot from the centrifuge tube to a 15-ml test tube. Store the aliquot at 4 C prior to further purification for HPLC analysis. It is recommended that analysis of extracts be conducted within 10 hr of extraction to minimize potential conversion of isoflavones. Glycitein 7,4 -Dihydroxy-6- methoxyisoflavone CH 3 Glycitin Acetylglycitin Malonylglycitin 7,4 -Dihydroxy-6- methoxyisoflavone-7-dglucoside 6 --Acetylglycitin 6 --Malonylglycitin H CH 3 H Me H C 2 H CH 3 H I1.6.3 Current Protocols in Food Analytical Chemistry

4 Table I1.6.1 Matrices of Some Common Soybean Products and Their Isoflavone Content Protein (% a ) Carbohydrate (% a ) Lipids (% a ) Isoflavones (mg/g on total basis b ) Soybeans c Full fat soy flour d Defatted soy flour e Soy milk powder f Soy containing bread g a Percent by weight on total basis. b Milligrams/gram on total basis from company certificate of analysis unless specified. c Harvested in hio. d Expeller Soy Flour I.P. from SunRich. e Defatted soy flour from Cargill Foods. f Soy Supreme Basic Soy milk powder from SunRich. g Soy-containing bread from Bavoy. Table I1.6.2 Solvent Mixture for Extracting Isoflavones from Soy Foods a Sample Acetonitrile 0.1 M HCl Water Dehydrated/solid sample 0.5 g 10 ml 2 ml 4 ml Liquid sample 2 ml 10 ml 2 ml 2 ml a Final mixture: 60% acetonitrile/40% water. Table I1.6.3 Commercial Sources of Soy Isoflavone Reference Standards a Isoflavone Apin Chemicals Fisher Scientific Indo Fine Chemical LC Laboratories Sigma Aldrich Daidzein X X X X X Daidzin X X X X Acetyldaidzin X Malonyldaidzin X Genistein X X X X X Genistin X X X X Acetylgenistin X Malonylgenistin X Glycitein X X X Glycitin X X X Acetylglycitin X Malonylglycitin X a rdering and product information can be found at the company websites (see SUPPLIERS APPENDIX). Analysis of Isoflavones in Soy Foods I1.6.4 Current Protocols in Food Analytical Chemistry

5 SAMPLE PREPARATIN FR EXTRACTIN Extract fresh food samples whenever possible. Grind solid and/or dehydrated soybean products in a grinder or with a mortar and pestle until a fine powder (one that can pass through a 50-mesh screen) is obtained. Mash moist soybean product to a fine paste. SUPPRT PRTCL 1 If food samples need to be stored for later extraction, seal in a plastic bag and store up to 10 days at 20 C or lower to prevent isoflavone loss and degradation. Thaw the sample completely at room temperature immediately before extraction. If freeze drying is desired, perform after food samples have been prefrozen in a freeze-dry flask to 20 C or lower. PREPARATIN F INDIVIDUAL ISFLAVNE STANDARDS In this protocol, commercially purchased isoflavone standards (Table I1.6.3) are dissolved in a suitable solvent to prepare solutions in a series of decreasing concentrations. The absorbances are then measured using a UV spectrophotometer set at the isoflavone s maximum wavelength (λ max ), and the concentrations of isoflavone standard solutions are calculated using published molar extinction coefficients. The spectrum is also scanned in order to evaluate the fine structure (see Background Information, Spectral fine structure). BASIC PRTCL 2 Materials Standard isoflavones (see Support Protocol 3 and Tables I1.6.3 and I1.6.4) 80% (v/v) methanol 5-digit analytical balance 10- and 50-ml volumetric flasks Ultrasonic bath 0.45-µm filter Analytical pipets Amber vials with caps UV spectrophotometer and quartz/glass cuvette Table I1.6.4 Isoflavone Stock Standard Solutions and Calibration Ranges Isoflavones Weight (mg) in 50 ml stock solution Stock solution (mg/ml) Working solutions (µg/ml) Daidzein Daidzin Acetyldaidzin Genistein Genistin Acetylgenistin Glycitein Glycitin Acetylglycitin I1.6.5 Current Protocols in Food Analytical Chemistry

6 Prepare stock standard solutions 1. Using a 5-digit analytical balance, weigh the amount of each crystalline isoflavone standard given in Table I1.6.4 and transfer to individual 50-ml volumetric flasks. 2. Add 80% methanol to near 50 ml, stopper each flask, and mix well by repeated inversion and by incubating in an ultrasonic bath until complete dissolution is achieved (~10 min). Dissolution may be aided by initial addition of a small amount of a more effective solvent (e.g., DMS) before bringing to volume for spectrophotometric measurement (no more than 1% of the total volume should be adequate). Dissolution can also be aided by using warm solvent ( 50 C). 3. Fill each flask with 80% methanol to the final volume (50 ml). Seal flasks tightly with a cap and store up to 6 months at 20 C. 4. Before use, completely thaw frozen stock solutions at room temperature and then pass through a 0.45-µm filter. Prepare working standard solutions 5. From each stock solution bottle, accurately transfer 0.1, 0.3, 0.5, 1, and 2 ml solution to 10-ml volumetric flasks using analytical pipets. 6. Dilute to volume using 80% methanol (see Table I1.6.4 for working concentrations). Mix well by repeated inversion of the flasks. 7. Transfer working solutions to tightly capped amber vials and store up to 2 months at 20 C. Completely thaw at room temperature before analysis. The working solutions will be used for UV absorbance measurement and for quantification by HPLC.Calibration should be performed at least every 2 months when new working standards are prepared. Measure absorbance using a UV spectrophotometer 8. Turn on the UV spectrophotometer and allow to warm up at least 30 minutes. 9. Zero the spectrophotometer with an appropriate blank (i.e., 80% methanol). Table I1.6.5 UV Spectral Pattern of Twelve Soy Isoflavones in 80:20 (v/v) Methanol/Water Analysis of Isoflavones in Soy Foods I1.6.6 Isoflavone MW Absorption peaks λ max (nm) ε λmax (liter/mol cm) a Daidzein Daidzin Acetyldaidzin Malonyldaidzin Genistein Genistin Acetylgenistin Malonylgenistin Glycitein Glycitin Acetylglycitin Malonylglycitin a Values from Murphy et al. (2002). Current Protocols in Food Analytical Chemistry

7 10. Fill a quartz/glass cuvette with the standard solution of a specific isoflavone. Measure the absorbance at λ max (see Table I1.6.5). Take reading immediately. With the dilution applied here, the absorbance reading should be in the range of 0 to bserve the UV spectrum (Table I1.6.5; also see Background Information, Spectral Fine Structure). 12. Repeat steps 9 to 11 for all standards and concentrations. Calculate isoflavone concentration 13. Calculate the concentration of isoflavone standard solutions based on the Beer-Lambert law. A = c (mol/liter) l (cm) ε (liter mol 1 cm 1 ) where A is absorbance, c is concentration, l is cell pathlength (typically 1 cm), and ε is the molar extinction coefficient (i.e., absorbance of a 1 M solution in a 1-cm light path at a specific wavelength; Table I1.6.5). GRADIENT SEPARATIN AND IDENTIFICATIN F ISFLAVNES USING A C18 REVERSED-PHASE CLUMN In this protocol, a reversed-phase separation using a C18 HPLC column and gradient mobile phase are used to separate isoflavones in the soy food extract. Reversed-phase HPLC operates on the basis of hydrophilicity and lipophilicity. Reversed-phase C18 columns consist of silica-based packings with covalently bound 18-alkyl chains. The mobile phase carries the sample solution through the column. The chemical and physical interactions of the mobile phase and sample with the column determine the degree of migration and separation of components in the column. In gradient elution, different isoflavone compounds are separated by increasing the strength of the organic solvent. Detection is achieved by monitoring UV absorbance at 260 nm using a UV or photodiode array (PDA) detector. Retention time, UV spectra, and co-elution are used to identify individual isoflavones in food extracts by comparison with pure isoflavone compounds (see Basic Protocol 2). BASIC PRTCL 3 Materials Solvent A: 1% (v/v) acetic acid in water, ph >2.0, fresh Solvent B: 100% acetonitrile Isoflavone reference standards (see Basic Protocol 2) Food sample: isoflavone extract from soy foods, after clean-up procedure (see Support Protocol 2) Sonicator, vacuum filtration device, or in-line vacuum degasser HPLC system: UV/Vis photodiode array (PDA) detector Pump: gradient Waters 2695 separation module with heating and cooling system Waters Nova-Pak C18 reversed-phase column ( mm, 4-µm i.d., 60-Å pore size) Waters Nova-Pak C18 guard column or a guard column with similar packing material Injector: automatic Computer data system Additional reagents and equipment for preparing calibration curves (see Basic Protocol 2) I1.6.7 Current Protocols in Food Analytical Chemistry

8 Prepare mobile phase 1. Check ph of solvent A, making sure it is >2.0. Degas both mobile phases via vacuum filtration, ultrasonic agitation, or inline vacuum degasser. Acetic acid serves as a modifier to protonate glucosides of isoflavones in the mobile phase. While freshly prepared solvent A is preferred, solvent that is up to 2 days old is acceptable. Set up HPLC apparatus 2. Turn on the UV/Vis PDA detector at least 30 minute before analysis. 3. Prime the gradient pump and flush the lines with mobile phases A and B, following manufacturer s instructions. 4. Program the mobile phase gradient as listed in Table I Set the following HPLC parameters: Pump flow rate: 0.6 ml/min Column oven temperature: 25 C Injection volume: 10 µl UV monitor: 260 nm. For strong UV absorbers such as isoflavones, total injected amounts of <20 ìg and injection volumes of 10 to 30 ìl are typical. Larger injection volumes can cause broadening of peaks. The flow rate was developed based on the column inner diameter and mass sensitivity. 6. Equilibrate the system and column with the starting mobile phase composition (i.e., 85% A:15% B) of mobile phase A for 10 column volumes ( ml/column volume), until the baseline is flat and smooth. The small particle size (4 ìm) of the bonded phases and small pore size (60 Å) of the C18 column efficiently resolve mixtures of similar isoflavone derivatives in the soy food matrix. Perform HPLC analysis 7. Inject each working calibration standard and standard mixture (including any optional internal standard) individually into the automatic injector. The standard mixture can be prepared by taking a known amount of each standard solution into a sample vial and mixing well. The concentration of each isoflavone in the standard mixture can be calculated with the dilution factor. To check the recovery of a chromatographic condition, it is common to spike a known concentration of standard solution into the sample solution. A matrix effect is suspected if the spike recovery is outside the limits of 90% to 110%. Recovery is calculated using the formula R = [(C s C)/S] 100, where R is percent recovery, C s is spiked sample concentration, C is sample background concentration, and S is concentration equivalent of spike added to sample. If a matrix effect is suspected, adjustment of chromatographic condition must take place. Table I1.6.6 Mobile Phase Gradients for Isoflavone Analysis Using a Reversed-Phase C18 Column Time (min) Solvent A (%) Solvent B (%) Analysis of Isoflavones in Soy Foods I Current Protocols in Food Analytical Chemistry

9 8. Measure the peak area of each working standard solution at 260 nm. Record the retention time for each peak. bserve the UV spectrum of each peak. Modern HPLC systems commonly include computer software that automatically integrates peak areas. For older systems, manual integration may be necessary. Besides the retention time, the UV spectra of each peak should be used to confirm the identity of isoflavones (see Background Information, Spectral Fine Structure). 9. Inject food sample and record data as in step 8. Analyze results 10. Prepare calibration curves by plotting peak area versus known concentration for each standard solution. The slope of the line will be used to calculate the concentration of the respective isoflavone. Peak area should be linearly proportional to analyte concentration (y = ax + b) with a correlation coefficient >0.98 and an intersect very near the origin. In this equation, y is the peak area, x is the concentration of standard solution, a is the response factor between peak area, and b is the concentration ( 0). 11. Identify isoflavones in the food sample by comparing the retention times of peaks with co-elution of isoflavone standards. Confirm identification by comparing the spectra of isoflavones in the food sample with those of standards (see Background Information). The complete separation of all twelve isoflavones using this chromatographic condition requires 30 min. The elution order is daidzin, glycitin, genistin, malonyldaidzin, malonylglycitin, acetyldaidzin, acetylglycitin, malonylgenistin, daidzein, glycitein, acetylgenistin, and genistein. Isoflavone retention and separation are influenced by column temperature. The oven temperature is set at 25 C. Absorbance (AU at 260 nm) %B Retention time (min) %B 35%B Figure I1.6.3 Gradient HPLC separation of isoflavone standards (see Basic Protocol 3). Peaks: 1, daidzin; 2, glycitin; 3, genistin; 4, malonyldaidzin; 5, malonylglycitin; 6, acetyldaidzin; 7, acetylglycitin; 8, malonylgenistin; 9, daidzein; 10, glycitein; 11, acetylgenistin; 12, genistein. Conditions: Waters Nova-Pak C18 reversed-phase column ( mm; 4-µm i.d.; 60 Å pore size); mobile phase: 1% acetic acid in water (solvent A) and acetonitrile (solvent B); flow rate: 0.60 ml/min; UV detector: 260 nm; column temperature: 25 C. The dotted line represents the gradient of solvent B. I1.6.9 Current Protocols in Food Analytical Chemistry

10 12. Use the area of each isoflavone peak and the calibration curves to calculate the isoflavone concentrations in the sample. Calculate the final concentration of isoflavones in soy food by considering the dilution factor during sample preparation. A sample chromatogram of isoflavone standards is shown in Figure I SUPPRT PRTCL 2 SAMPLE PREPARATIN FR HPLC ANALYSIS Soy foods are complex matrices (see Table I1.6.1). After initial aqueous extraction, the crude extract containing isoflavones is concentrated but may contain components that are immiscible with the HPLC mobile phase for direct injection. A clean-up step is necessary to prevent the introduction of particles and precipitates that may block the frit and foul the column. Materials Crude isoflavone extract (see Basic Protocol 1) Methanol 10-ml glass tube with cap Solvent evaporation apparatus (e.g., nitrogen gas, Speedvac) Ultrasonic bath 0.20-µm filter 1. Transfer 1 ml supernatant from crude soy extract to a 10-ml glass tube. Dry the aliquot under a nitrogen gas stream or using a Speedvac evaporator. 2. Resuspend the residue in 1 ml of 100% methanol. 3. Ultrasonicate the mixture in a bath for 10 min. This step is designed for complete protein precipitation, removal of methanol-insoluble sugars, and isoflavone redissolution in solvent. 4. Store the mixture up to 10 hr at 4 C if injection is not directly conducted. 5. Before injection, vortex the mixture 1 min and filter through a 0.20-µm filter. SUPPRT PRTCL 3 CNVERTING GLYCSIDIC ISFLAVNES T THEIR AGLYCNES USING ENZYMES Isoflavone aglycone standards are commercially available. However, the acetyl and malonyl glucosides of isoflavone standards are not readily available due to their lack of stability during transportation and storage. Hydrolysis of isoflavone glycosides removes the sugar moiety from the aglycones, thereby enabling quantification of total isoflavones in soy foods when the authentic standards of glycosidic isoflavones are not available. Hydrolysis is also a useful tool for structural analysis when specificity is achieved. β-glucosidase (Emulsin) and cellulase (from Aspergillus niger) have been used efficiently to convert isoflavone conjugates to their aglycone forms (Franke et al., 1994; Liggins et al., 1998). They remove the β-linked glucose from the 7-hydroxyl group on the isoflavone aglycones with adequate purity. Figure I1.6.4 illustrates the method using enzymatic hydrolysis for isoflavone analysis. Analysis of Isoflavones in Soy Foods By comparing the chromatogram from enzymatic hydrolysis with that in the absence of enzymatic hydrolysis, the completeness of hydrolysis by enzymes and the total isoflavones in soy food extract can be determined (see Basic Protocol 3). I Current Protocols in Food Analytical Chemistry

11 Materials Crude isoflavone extract (See Basic Protocol 1) Enzyme: β-glucosidase, almond-isolate or cellulase, Aspergillus niger extract 0.1 M ammonium acetate buffer, ph 5 (adjust ph with acetic acid) 100% methanol Solvent evaporation unit (nitrogen gas) Ultrasonic bath 0.2-µm filter 1. Centrifuge the crude isoflavone extract 30 min at 430 g, 25 C. 2. Take 1 ml supernatant from the crude extract and evaporate the solvent under nitrogen flow. 3. Dissolve 100 Fishman units of enzyme in 5 ml of 0.1 M ammonium acetate buffer, ph 5. ne Fishman unit is the amount of enzyme necessary to digest 1 ìmol substrate in 1 min at ph 5.0, 37 C. 4. Add 1 ml dissolved enzyme solution into the vial containing dried extract. Mix well in an ultrasonic bath for 10 min. 5. Incubate the solution 12 hr in a 37 C water bath. vernight is recommended. 6. Evaporate the solvent under nitrogen flow. 7. Dilute the dried residue with 1 ml of 100% methanol. Vortex and/or place in an ultrasonic bath to ensure solubilization. 8. Pass the sample solution through a 0.2-µm filter immediately before HPLC analysis. Solvent extraction of isoflavones from soy foods Evaporate solvent under nitrogen flow Dissolve dried residue with buffer containing enzymes Dissolve dried residue in 100% methanol Incubate at 37 C in a water bath for 12 hr Evaporate under nitrogen flow Redissolve dried residue in 100% methanol Filter before HPLC injection Figure I1.6.4 A flow diagram for isoflavone analysis using enzymatic hydrolysis. I Current Protocols in Food Analytical Chemistry

12 CMMENTARY Background Information Isoflavones Isoflavones comprise two benzene rings (A and B) linked through a heterocyclic pyrane C-ring at the 3 position (Fig. I1.6.1), which distinguishes them from flavones (Ingham, 1982; also see Figures I1.3.7 and I1.3.9 for comparison). The primary isoflavones in soybeans are the genistein, daidzein, and glycitein families (Fig. I1.6.2). Each family consists of its respective aglycone, β-glucoside, malonylglucoside, and acetylglucoside. ther isoflavones (i.e., formononetin or biochanin A) are present in only trace amounts (<1.0 µg/g). Malonylglucosides are the predominant form of isoflavones in unprocessed soybeans. Previous studies have concluded that malonylglucosides are heat labile and easily converted to their corresponding acetylglucosides and/or β-glucosides depending on the thermal conditions of processing and preparation (Xu et al., 2002). Enzyme activity may be responsible for the formation of aglycones by hydrolysis of glucosides (Coward et al., 1998). Various processing conditions produce soy products with a wide range of isoflavone content and composition. Recent studies observed that the chemical forms and abundance of isoflavones in soy foods have a significant impact on their bioavailability and biological effects (King, 1998; Izumi et al., 2000; Setchell et al., 2002). It is thus very important to avoid altering the natural forms and abundance of the twelve soy isoflavones during extraction, identification, and quantification. Extraction Soybeans and soybean products contain high levels of protein, carbohydrates, and lipids (Table I1.6.1). As minor components of complex mixtures, isoflavones must first be separated from the bulk of the matrix constituents prior to analysis. Efficient extraction methods for isoflavones should account for their diverse structures, chemical properties, and the food matrix of which they are constituents. This unit describes a practical way of extracting isoflavones from soybean products in their natural forms using readily available solvents and laboratory equipment. Before extracting a compound of interest from a complex food matrix, it is necessary to obtain knowledge about the physical and chemical nature of the sample i.e., moisture content, stability in acid and base, and thermal stability. The hydrophobicity of the isoflavone forms is aglycone > acetylglucoside > malonylglucoside > β-glucoside, based on their chromatographic behavior on reversed-phase columns in the presence of an acid in the mobile phase to protonate the glycosidic isoflavones. The ester bonds of acetyl- and malonylglucose of isoflavones are labile at elevated temperatures and under acidic or basic conditions. The aqueous solubilities of the isoflavone aglycones are low and are ph dependent due to the acidic nature of their phenolic groups. Conju AU nm Analysis of Isoflavones in Soy Foods I Figure I1.6.5 The UV characteristics of daidzein (dashed line), genistein (solid line), and glycitein (dotted line) in HPLC mobile phases (A: 1% acetic acid in water, B: 100% acetonitrile, A/B = 85:15) at 25 C monitored at 260 nm. Current Protocols in Food Analytical Chemistry

13 gation to glucose residues increases solubility, while acetylation or malonylation of the glucoses reduces solubility. Spectral fine structure of soy isoflavones The differences in the spectral characteristics of individual isoflavones are small but very important in their identification (Harborne, 1967). Isoflavones can be readily distinguished by their UV spectra, which typically exhibit an intense Band II (240 to 280 nm) absorption with only a shoulder or low-intensity peak representing Band I (300 to 330 nm). Band I absorption involves the B ring (cinnamoyl system; Figure I1.6.1), and Band II absorption involves the A ring (benzoyl system). The Band II absorption of isoflavones is relatively unaffected by increased hydroxylation of the B ring. Band II is, however, shifted bathochromically by increased oxygenation in the A ring. In addition to the absorption maxima of the isoflavones, the shape of the spectra provides important information for identification of pure isoflavone standards and isoflavones in soy extract. Fine structures of three soy isoflavone aglycones and their conjugates are shown in Figures I1.6.5 to I Details of the molecular characteristics of isoflavones can be found in The Chemistry of Flavonoid Compounds by llis (1962). Isoflavones in solution obey the Beer-Lambert law, where absorbance (A) equals concen AU nm genistein genistin acetylgenistin malonylgenistin Figure I1.6.6 UV spectra of the genistein family in HPLC mobile phases (A: 1% acetic acid in water, B: 100% acetonitrile, A/B = 85:15) at 25 C, monitored at 220 to 400 nm AU daidzein daidzin acetyldaidzin 0.00 malonyldaidzin nm Figure I1.6.7 UV spectra of the daidzein family in HPLC mobile phases (A: 1% acetic acid in water, B: 100% acetonitrile, A/B = 85:15) at 25 C monitored at 220 to 400 nm. I Current Protocols in Food Analytical Chemistry

14 Analysis of Isoflavones in Soy Foods I tration multiplied by the molar extinction coefficient (ε λmax ). The molar extinction coefficient is defined as the absorbance of a 1 M solution of isoflavones, in a defined solvent, in a 1-cmpathlength cuvette, at its maximum wavelength (λ max ). This information can be used to quantify the concentration of a certain isoflavone with known absorbance. Table I1.6.5 gives some known values of molar extinction coefficients of soy isoflavones. Variations in laboratory conditions (e.g., different solvents, temperature, absorbance wavelength) may be responsible for differences in reported extinction coefficients. The actual absorption and fine structure will depend on the composition of the mobile phase. A shift of 2 to 3 nm in the maximum absorption wavelength is usual. The spectrum of a specific isoflavone in soy food should be compared with an authentic pure standard. They should be identical for both the λ max and the fine structure. Isoflavone analysis using HPLC HPLC is the method of choice for the analysis of isoflavones in soy products. HPLC is fast, reproducible, requires small sample sizes, and can be used for both qualitative and quantitative analysis as well as for separation purposes. HPLC system. The basic HPLC system comprises a solvent delivery pump, injector, analytical column, and detector. Most isoflavone analyses are performed using a binary gradient, a revered-phase C18 column, and a UV detector. A photodiode array detector is capable of producing a UV/Vis absorption spectrum for each peak and has become the norm in isoflavone analysis. Isocratic elution was reported for isoflavone analysis but is not as effective as a gradient to separate the wide variety of isoflavones within an acceptable elution time. A typical gradient solvent used with reversed-phase C18 columns starts with a high proportion of polar solvents (i.e., water) and gradually increases the proportion of a less polar solvent (i.e., acetonitrile or methanol). The aqueous solvent is usually acidified to prevent ionization of isoflavone glycosides, which can give multiple peaks for some compounds. Identification. The order of elution of isoflavones is largely independent of minor variations in the solvent system, and thus it is possible to make tentative identifications by comparing relative retention times and co-elution with pure isoflavone standards. The actual retention times will vary between different runs, usually within 1 min. Such variation could be caused by differences in mobile phase preparation and the HPLC system (i.e., oven temperature, column condition). Quantification. For accurate quantification, a standard curve of peak area versus concentration should be constructed for each standard using the same chromatographic conditions (e.g., wavelength and solvent) as for the samples under analysis. The concentration range of standard curves should be determined according to both the isoflavone level of soy food samples and dilution factors during sample preparation such that the UV absorbance of the injected sample is within a range of 0 to 1. The appropriate standard curve can then be used to calculate the quantity of isoflavones represented by each HPLC peak in the sample. Internal standards. The primary function of an internal standard is in the determination of the reliability of extraction, sample preparation, and chromatographic procedure. An internal standard can be added to the original sample. By comparing the peak area of the internal standard in the chromatogram with that of a control, the losses due to sample preparation can be established. Alternatively, an internal standard can be added to the extraction solvent or to the sample prior to HPLC injection to measure the recovery rate of the chromatographic system. A suitable candidate for an internal standard for isoflavone analysis should (1) be similar in structure to isoflavones but not present in soy foods, and (2) be completely separable from isoflavones by the same chromatographic conditions at a specific wavelength. Apigenin, 2,4,4 -trihydroxydeoxybenzoin (THB), flavone, and equilenin have been reportedly used as internal standards. Hydrolysis The hydrolytic removal of the carbohydrate component of isoflavone glycosides simplifies the quantitative analysis of the isoflavones. Besides enzymatic hydrolysis (described in Support Protocol 3), acid hydrolysis and alkaline hydrolysis are also reportedly used. Acid hydrolysis is used primarily for cleaving sugars from glycosides, while alkaline hydrolysis finds application in the specific removal of acyl groups from acylated glycosides to produce β-glucosides of isoflavones (Klump et al., 2001). However, these methods are less commonly used due to a variety of reasons, including incomplete hydrolysis of the glycosides and degradation of the unconjugated isoflavones (Liggins et al., 1998). Current Protocols in Food Analytical Chemistry

15 Critical Parameters and Troubleshooting Extraction The solvents used for extracting isoflavones from soy foods were chosen according to the solubility of isoflavones and the food matrix involved. The diversity of soy isoflavones in polarity requires the use of a combination of organic solvent and water for extraction. The organic-to-water ratio (10:5) was established based on Murphy s study on solvent selection (Murphy et al., 2002). Water content in the solvent needs to be adjusted according to the moisture content in soy foods. Freeze-drying of samples before extraction simplifies the extraction process, but is not a prerequisite. Acetonitrile, acetone, ethanol, and methanol have been used to extract isoflavones from soy foods. Among them, acetonitrile proved to be the most efficient (Griffith et al., 2001; Murphy et al., 2002). The solvent is supplemented with 0.1 M HCl to completely un-ionize the isoflavones and to release them from protein complexes by denaturing and precipitating the proteins. Room temperature is recommended for extraction to avoid alteration of the natural forms of the isoflavones. The time for extraction, 2 hr, was chosen for maximum recovery and shortest processing time. Ultrasound is used to aid the extraction process by degrading and weakening the cellular matrix. Isoflavones are relatively stable compounds, but can degrade under certain conditions. Daidzein and genistein are light sensitive. It is necessary to work in dim light and to avoid exposure of extracts to air. Sample tubes should be capped during extraction. It is recommended to wrap sample flasks with aluminum foil to avoid light. Malonylglycosides of isoflavones are heat labile. Samples should be kept in a refrigerator between preparation and analysis, or kept in a freezer ( 20 C) for storage. The crude extract needs to be further purified for HPLC analysis. Direct injection of the crude extract into the HPLC would clog the frit and analytical column with precipitated impurities (i.e., proteins). Besides solvent extraction, supercritical carbon dioxide has been applied to extract isoflavones (Rostagno et al., 2002). Due to the hydrophobicity of carbon dioxide, supercritical fluid extraction is more suitable for extracting nonpolar aglycones than polar glycosides of isoflavones, and may not be quantitative. Standard solutions Follow good laboratory practices for UV spectrophotometry (UNIT F2.2). For the overall analysis, it is critical to prepare precise standard curves. Manufacturer s instructions and MSDS s should be considered to minimize hazards and avoid degradation prior to and during sample preparation. Pure isoflavone standards should be completely dissolved in solvent when preparing stock solutions. Ultrasound, addition of a small amount of a more effective solvent, and warming the solvent (up to 50 C) have proved to be efficient and safe methods for aiding dissolution of isoflavone standards without altering their original structure. Stock solutions should be kept at 20 C for storage purposes. The stock solution container needs to be sealed very well to avoid evaporation during storage. Stock solutions that have been stored frozen should be thawed at room temperature before reuse. Ultrasound is necessary to completely dissolve crystals that may develop during storage. All solutions should be filtered before reuse. Sample preparation Isoflavone glycosides can be deesterified and decarboxylated to their simpler conjugates under appropriate conditions. The conditions of sample handling and preparation prior to HPLC analysis are critical to ensure that there is no degradation or alteration of analytes. A handling temperature >60 C risks alteration of the original composition of isoflavones in the food matrix and should be avoided. Daylight or white fluorescent light should be avoided during sample preparation. Wrapping glass vials containing isoflavones in aluminum foil is suggested. When resuspending isoflavones in pure organic solvent, extra care should be taken to completely dissolve extracted isoflavones in the solvent. Solublization can be achieved by ultrasonification and providing sufficient time for the isoflavones to dissolve. Vortexing helps to release and recover isoflavones from precipitated proteins that may stick to the wall of glass vials and containers. All solvents must be of a high degree of purity. All extracted samples should be filtered to remove small particles. Hydrolysis Quantitative results for different foods using the hydrolysis method need to be treated with caution. Hydrolysis conditions should be adjusted based on the knowledge of a crude soy food extract i.e., the range of isoflavone con- I Current Protocols in Food Analytical Chemistry

16 Analysis of Isoflavones in Soy Foods I centration or the aglycone ratios. Partially hydrolyzed products lead to underestimation of the total isoflavones in the extracts. HPLC analysis Many parameters are critical to the successful and reproducible separation of isoflavones using HPLC. Backpressure Degassing should always be used to avoid outgassing and air bubble formation during HPLC analysis, especially when using a gradient. Backpressure can be used as an indicator for the condition of the column and the system. An abnormally high backpressure can be a sign of a fouled guard column. An abnormally low backpressure can be a sign of a leak between tubing and fittings, which causes poor peak shape. Variations in backpressure during analysis should be <50 psi. Larger variations could be caused by a pump malfunction. Column care Column equilibration (~10 column vol. recommended) ensures baseline stability, good peak shape, and reproducible retention times. For the specific column chosen in this unit for isoflavone analysis, at least 18 ml total is needed to equilibrate the system with mobile phase. The common practice is to purge the pump system and connect the inlet end of the column to the injector outlet. The initial pump flow should be set at 0.1 ml/min and increased to 0.6 ml/min in 0.1 ml/min increments. nce a steady backpressure and baseline have been achieved, the column is ready to use. Before injecting samples, it is suggested to run a blank gradient first to clean the column and help check for the possibility of impurity peaks. For overnight storage, the column should be kept flushing at 0.1 ml/min with a mild solvent (i.e., acetonitrile or methanol). This practice reduces the re-equilibration time the following day to a few minutes rather than an hour or more. For long-term storage, the column should be flushed with its shipping solvent (i.e., acetonitrile for a Waters Nova-Pak RP-C18 column) at 1 ml/min for at least 15 min before being taken off and tightly sealed at both ends. When contamination collects on a column, poor peak shapes (usually tailing) and extraneous peaks will appear with an increase in backpressure. Therefore, samples should always be filtered before injection. Mobile phases should be filtered a few days after preparation to remove any particles and prevent microbial growth. Nevertheless, after repeated use, the column needs to be cleaned or regenerated, because tightly bound impurities will bind to the stationary phase and reduce the attraction the column packing has for the sample components, thus reducing the retention times for the peaks of interest as the column ages. The manufacturer s instructions should be followed to regenerate the column. To prolong the life of expensive analytical columns, it is recommended that a filter and a guard column be placed before the analytical column to serve as a protective factor. Solvent preparation To prepare mobile phases, clean HPLCgrade solvent must be used along with clean solvent bottles. Since most of the organic solvents are volatile, they should be well sealed in solvent bottles after preparation. Acetic acid is subject to evaporation, causing the ph of the aqueous acetic acid solution to shift. Freshly prepared acetic acid in water is recommended for each set of analyses. The calculated ph of 1% acetic acid in water is 2.7; however, it is suggested to check the ph of the acid solution each time it is prepared to assure its ph is >2. Anticipated Results The extraction protocol typically recovers 96% to 108% of isoflavones. The HPLC protocol typically recovers isoflavones in a range between 97% and 105% when tested using spiked reference standards. During reversedphase HPLC separation, all isoflavones elute from the column in the order of polar to nonpolar and are separated within 50 min of run time. Time Considerations When freeze-drying is not included, completion of the full extraction process requires 1/2 day. It is recommended that freeze-drying, which requires at least 12 hr, be performed overnight before the day of extraction. Calibration prior to HPLC requires 1 or 2 days to perform all necessary steps (i.e., prepare stock solutions, filter, measure absorbance, dilute, and HPLC injection of the working solutions). nce the stock solution and working solutions have been prepared, a calibration check can be performed on a UV spectrophotometer within 1 hr. The sample clean-up procedure for HPLC analysis may require a few hours. Enzymatic hydrolysis requires overnight incubation. Each HPLC analysis requires 50 min. When an autosampler is available, Current Protocols in Food Analytical Chemistry

17 samples can be injected and analyzed overnight. Typically, one technician can extract 12 samples per day to be analyzed automatically overnight. Literature Cited Birt, D.F., Hendrich, S., and Wang, W Dietary agents in cancer prevention: Flavonoids and isoflavonoids. Pharmacol. Ther. 90: Coward, L., Barnes, N.C., Setchell, K.D.R., and Barnes, S Genistein, daidzein, and their β-glucoside conjugates: Antitumor isoflavones in soybean foods from American and Asian diets. J. Agric. Food Chem. 41: Coward, L., Smith, M., Kirk, M., and Barnes, S Chemical modification of isoflavones in soyfoods during cooking and processing. Am. J. Clin. Nutr. 68:1486S-1491S. Fournier, D.B., Erdman, J.W., and Gordon, G.B Soy, its components, and cancer prevention: A review of the in vitro, animal and human data. Cancer Epidemiol. Biomarkers Prev. 7: Franke, A., Custer, L.J., Carmencita, M.C., and Narala, K.K Quantitation of phytoestrogens in legumes by HPLC. J. Agric. Food Chem. 42: Griffth, A.P. and Collison, M.W Improved methods for the extraction and analysis of isoflavones from soy-containing foods and nutritional supplements by reversed-phase high-performance liquid chromatography and liquid chromatography-mass spectrometry. J. Chromatog. A 913: Harborne, J.B Isoflavones. In Comparative Biochemistry of the Flavonoids, pp Academic Press, London. Ingham, J.L Phytoalexins from the Leguminosae. In Phytoalexins (Bailey and Mansfield, eds.) pp John Wiley & Sons, New York. Izumi, T., Piskula, M.K., sawa, S., bata, A., Tobe, K., Saito, M., Kataoka, S., Kubota, Y., and Kikuchi, M Soy isoflavone aglycones are absorbed faster and in higher amounts than their glucosides in human. J. Nutr. 130: King, R.A Daidzein conjugates are more bioavailable than genistein conjugates in rats. Am. J. Clin. Nutr. 68(Suppl):1496S-1499S Klump, S., Allred, M., MacDonald, J., and Ballam, J Determination of isoflavones in soy and select foods containing soy by extraction, saponification, and LC. Collaborative study. J. AAC Int. 84: Liggins, J., Bluck, L.J.C., Coward, W.A., and Bingham, S.A Extraction and quantification of daidzein and genistein in foods. Anal. Biochem. 264:1-7. Murphy, P.A., Barua, K., and Hauck, C.C Solvent extraction selection in the determination of isoflavones in soy foods. J. Chromatog. B 777: Nguyenle, T., Wang, E., and Cheung, A.P An investigation on the extraction and concentration of isoflavones in soy-based products. J. Pharm. Biomed. Anal. 14: llis, W.D The isoflavones. In The Chemistry of Flavonoid Compounds (T.A. Geissman, ed.) pp MacMillan, New York. Rostagno, M.A., Arafajo, J.M.A., and Sandi, D Supercritical fluid extraction of isoflavones from soybean flour. Food Chem. 78: Setchell, K.D Phytoestrogens: The biochemistry, physiology, and implications for human health of soy isoflavones. Am. J. Clin. Nutr. 68(Suppl.):1333S-1346S. Setchell, K.D. and Cassidy, A Dietary isoflavones: Biological effects and relevance to human health. J. Nutr. 129:758S-767S. Setchell, K.D., Brown, N.M., Zimmer-Nechemias, L., Brashear, W.T., Wolfe, B.E., Kirschner, A.S., and Heubi, J.E Evidence for lack of absorption of soy isoflavone glycosides in humans, supporting the crucial role of intestinal metabolism for bioavailability. Am. J. Clin. Nutr. 76: Xu, Z., Wu, Q., and Godber, S Stabilities of daidzin, glycitin, genistin, and generation of derivatives during heating. J. Agric. Food Chem. 50: Key References Griffith and Collison, See above. A systematic investigation in extraction and analysis of isoflavones from soy-containing foods and nutritional supplements. Murphy et al., See above. A systematic review and comparison of different solvents and aqueous-to-solvent ratios for isoflavone extraction efficiency. Contributed by Yu Chu Zhang and Steven J. Schwartz The hio State University Columbus, hio I Current Protocols in Food Analytical Chemistry