Chapter 2. Isoflavones: Source and Metabolism

6409_book.fm Page 24 Saturday, September 16, 2006 9:54 AM



24



Handbook of Nutraceuticals and Functional Foods



OH

R2

H OR3



O



R1

H



O



HO

O



HO

H

H



O



OH

H



Isoflavone

G

Gl

D

MG

MG l

MD

AG

AGl

AD



R1



R2



R3



H

OCH3

H

H

OCH3

H

HO

OCH3

H



OH

H

H

OH

H

H

H

H

H



H

H

H

OCCH2COOH

OCCH2COOH

OCCH2COOH

COCH3

COCH3

COCH3



HO

O



R2

R1



O



OH



Isoflavone



R1



R2



Daidzein

Genistein

Glycitein



H

H

OCH3



H

OH

H



FIGURE 2.1 Isoflavone glucoside and aglucon structures.



in the literature is improving as more authors are using the correct nomenclature in presenting

their isoflavone data. A summary of method reports in Table 2.1 reveal 34% of the reports used

μmole per g food and 6% reported isoflavones as aglucon totals. However, 60% of the papers cited

in this table are still using confusing and incorrect concentration units to express their data. Thus,

authors, reviewers and journal editors still need to rectify the problems and misinterpretation in

the scientific literature.

One of the reasons for the problems in accurate reporting of isoflavone data in the literature

results from the methodology used to analyze isoflavones in food matrices. Unfortunately, the

analytical quality control used by laboratories evaluating isoflavones is not even. Not all researchers

account for all 12 of the isoflavone forms found in most foods. Too many authors use only aglucon

standards to quantify all 12 forms. The extraction protocols reported in the literature continue to

use solvents that have been demonstrated to underestimate several of the isoflavone forms. Table



Dein, Gein



Dein, Gein, Glein,

D, G, Gl, MD, MG,

MGl



Dein, Gein, Glein,

D, G, Gl, MD, MG,

MGl



None reported



None reported



Dein, Gein



Dein, Gein, Glein,

D, G, Gl



HPLC-UV DAD

C18, 250 × 4.5 mm

Gradient



HPLC-UV DAD

C18, 250 × 4.5 mm

Gradient



HPLC-UV DAD and

ESI-MS

C18, 100 × 4.6 mm

Gradient, 4 mL/min



HPLC-UV DAD and

tandem-MS

C18, 250 × 4.5 mm

Gradient



HPLC-UV 254 nm

C18, 250 × 4.5 mm

Step Gradient



HPLC-UV DAD

C18, 250 × 4.5 mm

Gradient



Isoflavones

Reporteda



ELISA



Method



TABLE 2.1

Food Analysis of Isoflavones



80% hot

MeOH



acid

hydrolysis



80% MeOH

glucosides by

alkaline hydrolysis



20% DMSO in MeOH

sonication,

time unknown



80% MeOH



80% ACN or

80% MeOH or

80% EtOH



Glucuronidase hydrolysis



Sample Extraction



Gein



unknown



Dein, Gein



Dein, Gein, D, G



Dein, Gein, Glein, D,

G, Gl



Dein, Gein, Glein, D,

G, Gl



Dein, Gein



Standards Used



E



E



FL



Internal

Standardb



101



101



Recovery

(%)



9



μg/g



Source



Continued.



10



8



μg/g?

MW adj unknown



mg/100g



7



%



6



5



μg/g individual forms

no MW adj



mg/l as aglucons



4



mg/tablet, μg/m



Data Reported As



6409_book.fm Page 25 Saturday, September 16, 2006 9:54 AM



Isoflavones: Source and Metabolism

25



Dein, Gein,

Glein, D, G, Gl



Dein, Gein,

Glein, D, G, Gl

MD, MG, MGl,

AD, AG, AGl



Dein, Gein,

Glein, D, G, Gl

MD, MG, MGl,

AD, AG, AGl



Dein, Gein



Dein, Gein,

Glein, D, G, Gl



HPLC-UV

C18, 250 × 5 mm

Gradient



HPLC-UV

C18, 250 × 5 mm

Gradient



HPLC-UV

C18, 250 × 5 mm

Gradient



CE-ampermetric

Detector



HPLC-UV CEAD

C18, 250 × 4.5 mm

Gradient



Dein, Gein



83% MeOH

or 80% EtOH

2M NaOH

65oC



70% EtOH



80% MeOH



80% MeOH



80% MeOH



90% MeOH



Sample Extraction



80% MeOH



Dein, Gein,

D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein



G, D,

Gein, Dein



G, D

Gein, Dein



G, D,

Gein, Dein



Dein, Gein,

D, G



Standards Used



Dein, Gein,

Glein, D, G, Gl



Dein, Gein



FL



FL

added to

extract



Internal

Standardb



100



96



103-106



Recovery

(%)



11, 12, 20



13



14



15



16



17, 31



18



μmol/mL



μmol/g



μmol/g



μg/g

μg/g individual forms

μg/g adj MW totals



μg/l aglucons



Source



μg/g

no MW adj



Data Reported As



26



HPLC-MS-MS

C18, 250 × 5 mm

Gradient



Dein, Gein

D, G



Isoflavones

Reporteda



HPLC-UV DAD and

ESI-MS

C18, 250 × 4.6 mm

Gradient



Method



TABLE 2.1 (Continued)

Food Analysis of Isoflavones



6409_book.fm Page 26 Saturday, September 16, 2006 9:54 AM



Handbook of Nutraceuticals and Functional Foods



Gein,

G,

MG, AG



Dein, Gein, Glein,

D, G, Gl, MD, MG,

MGl



Dein, Gein,

Glein, D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein,

Glein, D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein,

Glein, D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein,

Glein, D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein,

Glein, D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein,

Glein, D, G, Gl



HPLC-UV

C18, 250 × 5 mm

Gradient



HPLC-UV DAD

C18, 250 × 4.5 mm

Gradient



HPLC-UV DAD

C18, 250 × 4.5 mm

Gradient



HPLC-UV DAD

C18, 250 × 4.5 mm

Gradient



HPLC-UV DAD

C18, 250 × 4.5 mm

Gradient



HPLC-UV DAD

C18, 250 × 3 mm

Gradient



HPLC-UV DAD

Column not specified

Gradient



HPLC-UV DAD

C18, 250 × 4.6 mm

Gradient

83% MeOH

2M NaOH

65oC



50% acetone

0.1 N HCl



50% ACN



83% ACN



80% MeOH



83% ACN



80% MeOH



80% MeOH



Dein, Gein,

D, G, Gl



Dein, Gein,

D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein,

D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein,

D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



unclear



Dein, Gein,

D, G, Gl,

MD, MG, MGl,

AD, AG



Dein, Gein,

Glein, D, G, Gl



G, AG, MG

Gein



Dein, Gein,

Glein, D, G, Gl



FO added

to extract;

all 12

isoflavones

added to food



D, G,

Gl, THB



D, G,

Gl



D, G,

Gl



65-92



75-100



81-98



81-98



Isoflavones: Source and Metabolism



Continued.



27



μmol/g



24, 36,

40



μmol/g



26



23



μg/g individual forms

μg/g no MW adj totals



μg/g individual forms

μg/g no MW adj totals



22



μg/g individual forms

μg/g no MW adj totals



25



21



μmol/g



no data



19



μg/ml



6409_book.fm Page 27 Saturday, September 16, 2006 9:54 AM



27



Dein, Gein,

D, G



Dein, Gein,

Glein



Dein, Gein,

Glein, D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein,

Glein, D, G,

Gl, AD, AG,

AGl, MD, MG

MGl



Gein, G



Dein, Gein,

Glein, D, G,

Gl, AD, AG,

AGl, MD, MG,

MGl



HPLC-UV DAD

ESI-MS

C18, 250 × 405 mm

Gradient



HPLC-UV DAD

C18, 150 × 3.9 mm

Gradient



HPLC-UV DAD

C18 250 × 4 mm

Gradient



HPLC-UV DAD

C18, 250 × 5 mm

Gradient



HPLC-UV DAD

C18 150 × 3.9 mm

Gradient



Isoflavones

Reporteda



HPLC-UV DAD

C18, 250 × 4.5 mm

Gradient



Method



TABLE 2.1 (Continued)

Food Analysis of Isoflavones



83% ACN

+ acid



80% MeOH

60oC



specifics

not described



Gein, G



Dein, Gein,

Glein, G, D

Gl



Dein, Gein,

D, G, Gl,

MD, MG, MGl,

AD, AG



Dein, Gein

Glein



D, G, Gein

Dein



Standards Used



equilenin

added to

extract



FL, Glein



FL



Internal

Standardb



91



72-94



98, 97



Recovery

(%)



30



32



33



μmol/g



μg/g individual

μg/g total no MW adj



μg/g, no MW adj



34



29



μg/g individual forms

μg/g no MW adj totals



nmol/g



28



Source



μg/g individual forms

μg/g no MW adj totals



Data Reported As



28



80% MeOH



83% ACN



80% MeOH

& acid

hydrolysis



80% MeOH



Sample Extraction



6409_book.fm Page 28 Saturday, September 16, 2006 9:54 AM



Handbook of Nutraceuticals and Functional Foods



80% EtOH



90% MeOH



Dein, Gein,

Glein, D, G, Gl



Dein, Gein,

Glein, D, G, Gl



aglucons,

β-glucosides,

malonyls,

acetyls



Dein, Gein,

Glein, D, G,

Gl, AD, AG,

AGl, MD, MG,

MGl



Dein, Gein,

Glein, D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein,

Glein, D, G, Gl



HPLC-UV DAD

C18, 250 × 4.5 mm

Gradient



HPLC-UV DAD

C18, 150 × 4.5 mm

Gradient



HPLC-UV DAD

C18 250 × 4 mm

Gradient



HPLC-UV DAD

& MS

C18 250 × 3.2

Gradient



HPLC-UV DAD

C18, 250 × 4.5 mm

Gradient



HPLC-UV DAD

C18, 150 × 4.5 mm

Gradient

75% EtOH



70% EtOH

reported



80% MeOH

ASE



83% MeOH

2M NaOH

65oC



80% MeOH



Dein, Gein,

Glein



HPLC-UV DAD

C18 150 × 4 mm

Gradient



80% MeOH



Dein, Gein,

Glein, D, G



HPLC-UV DAD

C18 250 × 4 mm

Gradient



38



μmol/g individual

na total



Dein, Gein, Glein

D, G, Gl



unknown



mole%



c



Dein, Gein,

D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



mg/kg individual

mg/kg MW adj



T



37



μmol/g

μg/g total no MW adj



Continued.



43



42



41



39



37



μg/g individual

μg/g total no MW adj



%

? MW adj



35



mg/g individual

mg/g total no MW adj



Gein, Dein, Glein

D, G, Gl, FO, B, C



none



Gein,

D, G, Gl,

FO, B



Dein, Gein,

D, G, Gl



Dein, Gein, Glein,

D, G, Gl



Dein, Gein,



6409_book.fm Page 29 Saturday, September 16, 2006 9:54 AM



Isoflavones: Source and Metabolism

29



solvent



diff% MeOH

or ACN + H+



Dein and glucosylated

products



Dein, Gein,

Glein, D, G,

Gl, AD, AG,

AGl, MD, MG,

MGl



HPLC-UV

ESI-MS

C18, 150 × 4.5 mm

Gradient



HPLC-UV DAD

C18

Gradient



47



μmol/g



45



μg/g individual forms

and %



Gein, G

D, AG



44



Source

μg/l individual forms



Data Reported As



46



not added

to sample



99



87, 62 52

in H2O



Recovery

(%)



nM



FL, Gein, G

D, AG



Dein, Gein,

D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein, G



Internal

Standardb



Dein



Dein, Gein,

D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



Dein, Gein

Glein



Standards Used



30



50% EtOH

60oC



Dein, Gein,

Glein, D, G, Gl,

MD, MG, MGl,

AD, AG, AGl



HPLC-UV DAD

C18, 100 × 4.6 mm

Gradient



SPE of

waste H2O



Sample Extraction



Dein, Gein,

G



Isoflavones

Reporteda



HPLC-UV DAD

ESI-MS

C18, 150 × 4.5 mm

Gradient



Method



TABLE 2.1 (Continued)

Food Analysis of Isoflavones



6409_book.fm Page 30 Saturday, September 16, 2006 9:54 AM



Handbook of Nutraceuticals and Functional Foods



Dein, Gein,

Glein, D, G, Gl,

MD, MG, MGl



Dein, Gein,

Glein, D, G, Gl

MD, MG, MGl



Gein, G, D



HPLC-UV DAD

EPI-MS

C18, 100 × 4.6 mm

Gradient



HPLC-UV

C18, 300 × 4.6 mm

Gradient



HPLC-UV 365 nm

C18, 300 × 4.6 mm

Gradient

80% ACN or

acid hydrolysis



80% MeOH



50% EtOH

60oC



53% ACN



none listed



Dein, Gein, Glein,

D, G, Gl



Dein, Gein,

D, G, Gl,

MD, MG, MGl



none listed

BA, added

to extract



50



51



μg/ml



49



μM



mg aglucon/100 g

mole%



48



μmol/g



Dein = daidzein, Gein = genistein, Glein = glycitein, D = daidzin, G = genistin, Gl = glycitin, MD = 6″-O-malonyldaidzin, MG = 6″-O-malonylgenistin, MGl = 6″-O-malonylglycitein,

AD = 6″-O-acetyldaidzin, AG = 6″-O-acetylgenistin, AGl = 6″-O-acetylglycitin, F = fluorescein, THB = 2,4,4′-trihydroxydeoxybenzoin, FL = flavone, FO = formononetin, B = biochanin

A; C = coumesterol, T = theophylline, BA = benzoic acid, adj MW = weight data adjusted for molecular weight differences of isoflavone glucosides and aglucons, DAD = photodiode

array detection, MS-SIM = mass spectrum-single ion monitoring, EC = electrochemical, CE = capillary electrophoresis, CEAD = coulometric electron array detector, ASE = accelerated

solvent extractor, na = not applicable.

b Internal standard and/or recovery spike added to dry food matrix before extraction solvents unless noted otherwise.

c MDin and MGin not available from reported supplier (Fisher Scientific).



a



Dein, Gein,

Glein, D, G, Gl,

MD, MG, MGl



HPLC-UV

C18, 250 × 4.5 mm

Gradient



6409_book.fm Page 31 Saturday, September 16, 2006 9:54 AM



Isoflavones: Source and Metabolism

31



6409_book.fm Page 32 Saturday, September 16, 2006 9:54 AM



32



Handbook of Nutraceuticals and Functional Foods



2.1 tabulates the analytical quality control used by 47 isoflavone analysis reports in the literature

since Murphy (2004).1

The current USDA-Iowa State University Isoflavone Database 1.3 contains data on 128 foods

and is currently undergoing an update by the USDA Food Composition Laboratory.4 These data

are derived for the peer-reviewed literature.



II. ANALYSIS METHODS

The methodology for HPLC isoflavone analysis continues to be refined. Bennetau-Pelissero et

al. reported an ELISA method for the aglucons.5 Food and supplement extracts were analyzed

after enzymatic hydrolysis with glucoronidase/sulfatase, apparently, although no data were presented on the efficacy of these enzymes hydrolyzing glucosides. Antonelli et al. reported total

isoflavones after alkaline hydrolysis of glucosides.6 Peng et al. reported using capillary electrophoresis with electrochemical detection for only genistein and daidzein from 70% ethanol sonicated extracts of foods.7 Several “fast” HPLC procedures have been reported with flow rates > 2

ml/min and short columns (10 cm).8–10 Apers et al. report good analytical quality control (QC)

data but give no data on concentrations in food or other biological samples.8 Klejdus et al. report

on accelerated solvent extraction (ACE) protocols for isoflavones from soy.9–11 Klejdus et al.

compared sonication at room temperature, Soxhlet extraction at solvent boiling point and ACE

at solvent boiling point, all in 90% aqueous methanol.11 Very low yields of extraction amounts

(by a factor of 10) were reported compared to their room temperature extractions. Klejdus et al.

report some of the same data.9 Klejdus et al. report using ACE conditions of 140oC and 140 bar

for a few food samples.10 Malonyl-β-glucosides concentrations are not reported but are clearly

apparent in their chromatograms. Rostango et al. reported on a solid phase extraction technique

for soy isoflavone analysis using initial 50% ethanol extraction at 60oC for 30 min and data

reported in μg/g for all 12 forms.12 Rostango et al. evaluated the stability of isoflavone extracts

and reported data in μM.12 Sample extracts stored in ethanol between –20 and 10oC were stable

for one week. Lin and Guisti thoroughly evaluated 83% acetonitrile vs. 53% acetonitrile ± acid

for isoflavone extraction but only for soybeans, not other foods.13 Kawanishi et al. measured the

isoflavone levels in waste water effluent in Osaka and reported averages of 143 μg/l genistein

and 43 μg/l daidzein.14



A. ISOFLAVONES



IN



SOY INGREDIENTS (INCLUDING SOYBEANS)



There are about 24 reports documenting the isoflavone levels in soy ingredients. These ingredients,

which consumers rarely have direct access to, are soybeans, defatted soy flours, soy protein

concentrates, soy protein isolates, and texturized vegetable protein. Most of these reports extracted

the isoflavones with 80% methanol, report their data in μg isoflavone/g ingredient, do not adjust

for the difference in molecular weight of the different isoflavone forms, and use a limited number

of authentic isoflavone standards for quantification. However, there have been a few reports that

give us accurate insight into isoflavone levels in ingredients since the 2004 review.

Achouri et al. reported isoflavone levels in defatted soy flours and soy protein isolates that were

lower than literature values.15 The solvent volume to sample weight ratio was 5 or less and was

probably the reason for the low analytical values. Although isoflavones are soluble in alcoholic

solvents, their solubility is limited, even in the best solvents. A minimum of 10:1 solvent to weight

is needed with an even larger ratio for samples very concentrated in isoflavones such as soy germ.

Park et al. reported on isoflavone levels in the anticarcinogenic protein lunasin but provided

no details on how analysis was performed.16 These authors stated that isoflavones levels in mature

soybeans were lower than immature beans in contrast to all other reports comparing maturity.17

Hubert et al. compared isoflavone with soyasaponin μmole levels in soybean seeds, soy germ, and

soy supplements.18 The isoflavone/saponin were 1 to 3 for soybeans and soy germ but ranged from



6409_book.fm Page 33 Saturday, September 16, 2006 9:54 AM



Isoflavones: Source and Metabolism



33



2–17 in soy supplements. Lee et al. report levels of isoflavones in 2002 Ohio-grown soybeans using

good analytical QC.19 The isoflavone contents ranged from 4.2 to 11.8 μmole/g.

Yu et al. attempted altering the synthetic pathway of isoflavone synthesis in soybeans to alter

distribution of genistein and daidzein forms as well as total isoflavone synthesis.20 Two analytical

methods were used for different soybean generations. All 12 isoflavones were reported to be

analyzed in F1 and F2 generations, whereas the AOAC method, which yields results for 6 isoflavone

forms (the other six are converted to β-glucosides), was used for the F3 generation.21 Samples were

all extracted with 80% methanol that under-extracts total isoflavones in this type of matrix.22 Duke

et al. evaluated isoflavone levels in glyphosate-treated glyphosate-resistant soybeans using an

unusual combination of analytical methods.23 Soy flours were extracted using accelerated solvent

extraction technology with 80% methanol, but no validation data were provided to compare with

conventional extract procedures. The β-glucosides and genistein were quantified by HPLC and

daidzein and glycitein were quantified by gas chromatography. The concentrations reported the

aglucons are very high compared to other reports in literature for control as well as experimental

treatments. Li et al. describe a procedure to glucosylate daidzein on multiple hydroxyl groups using

Thermotoga maritima maltosyltransferase.24 They created daidzein forms that were very hydrophilic. The authors quantified their products by HPLC using a daidzein standard curve, which was

reasonable because they were using pure daidzein in this model system.

Variyov et al. reported the dose effect of irradiation on isoflavones in soybeans.25 When

isoflavone contents are recalculated on a μmol/g basis, the glucoside content decreases markedly

with irradiation dose from 3.4 μmol/g to 1.8 μmol/g; however, the aglucon contents remained

unchanged averaging 0.26 ± 0.03 μmol/g. Unfortunately, the authors used a hot methanol extraction

method, thus giving a very limited picture of the processing effects.

Barbosa et al. reported the distribution of isoflavones in soy protein isolate manufacture on a

laboratory scale.26 These authors use an unusual polyamide column preparation of the hot 80%

methanol extracts. They report acidified water washing compared to water washing resulted in soy

protein isolates with higher isoflavone concentrations. Fukui et al. described a chromatography

process using a hydrophobic Diaion HP20 column to produce an isoflavone-free soy protein

isolate.27 This isoflavone-free protein isolate reduced serum cholesterol in a rat feeding study.

Rickert et al. reported mass balance distributions of isoflavones and saponins during pilot plant

manufacture of the isolated soy proteins, glycinin and β-conglycinin, as well as an intermediate

protein product.28 At the laboratory scale, increased temperature of protein extraction, and to a

limited extent, solvent to flake ratio, resulted in more isoflavones in the intermediate protein fraction,

a more denatured protein than either the glycinin or β-conglycinin fractions. However, these altered

distributions did not translate to 10 kg defatted soy flake pilot plant scale extraction.

Rickert et al. demonstrated that increasing temperature and pH significantly increases saponin and

isoflavone concentrations during soy protein isolate manufacture.29 These authors show that initial pH

neutralization in analytical extraction of isoflavones and saponins from soy protein isolates increased

recovery rate and normalized the mass balance discrepancies observed without pH adjustment.

Xu et al. report a carefully performed study to determine effectiveness in using reverse-osmosis

and ultrafiltration to fractionate isoflavones from soymilk and to further concentrate the isoflavone

permeate.30 The mass balance analysis revealed about 50% of mole mass of isoflavones could be

recovered in retentate for other uses such as supplements or food additives. These authors suggest

that additional optimization would increase yield.

Penalvo et al. evaluated commercial soy supplements and reported all but one had isoflavone

levels 1–76% lower than label claim.31 Penalvo et al.32 reported on the analytical quality control

of their modification of the Klump et al. method21 by using an electrochemical detector. Chua et

al. examined 13 reportedly isoflavone-containing supplements.33 Unfortunately, the authors

extracted their samples in 75% ethanol at room temperature for 24 h, conditions that under-extract

isoflavones from most matrices.22 The authors seem totally unaware of the acetyl- and malonylisoflavone forms and report these as impurities in their chromatograms.



6409_book.fm Page 34 Saturday, September 16, 2006 9:54 AM



34



Handbook of Nutraceuticals and Functional Foods



Pinto et al. conducted isoflavone distribution storage studies at –18 to 42oC and at several water

activities at 40oC for defatted soy flour and soy protein isolates.34 The temperature storage study

revealed no change in total mole mass of isoflavone at all temperatures. However, at 42oC, 10–20%

of the malonyl-glucosides were converted to their β-glucosides. These conversions were attributed

to the effects of temperature. Under controlled water activity conditions, major conversions of

glucosides to aglucons occurred in soy flours, but not soy protein isolates, at Aw = 0.87 but not at

lower water activities. These conversions were attributed to the effect of native glucosidases in soy

flour. Kim et al. compared isoflavone levels for 8 Korean soybean varieties after 3 y storage without

adjusting for molecular weight differences of the isoflavone forms.35 Under room temperature

storage, malonyl-β-glucosides decreased by a factor of 3 on mole basis but other isoflavone forms

remained unchanged. At –30oC, extractable malonyl-β-glucosides decreased and β-glucosides

increased but not on an equimolar basis. The authors do not indicate if data are on an “as is” or

dry weight basis.

Ismail and Hayes investigated the effects of β-glucosidase action on the different isoflavone

glucoside standards.36 The authors report using all 12 isoflavone standards but cited Fisher Scientific

as a source for malonyl-genistin and malonyl-daidzin, which cannot be correct. The data are reported

in mole%. These authors report no change in the concentrations of acetyl-genistein and malonylgenistin incubated 2 h at 37oC at pH 2, which seems unusual. The β-glucosides showed mixed

specificity for the isoflavone forms. Mixtures of isoflavone glucosides were hydrolyzed faster than

single glucosides. Almond β-glucoside hydrolyzed less than E. coli β-glucosidase. The E. coli βglucosidase preferentially hydrolyzed the β-glucosides with far less activity for the malonyl-βglucosides and acetyl-β-glucoside in the order of 90:6:5 mole% and these activities were consistent

for daidzein, genistein and glycitein forms. Whether this specificity holds for food matrices containing isoflavones remains to be examined.

Choi et al. examined the effect of several strains of lactic acid bacteria’s β-glucosidases to

hydrolyze isoflavone glucosides.37 Yields of 70–80% for genistin and 25–40% for daidzin were

reported. However, some lactic acid bacteria strains did not hydrolyze any isoflavone glucosides.

Isoflavones were analyzed as aglucons and amount hydrolysis was calculated by difference.



B. ISOFLAVONES



IN



SOY FOODS



Isoflavones in soy foods were reported in about 14 citations. The same analytical and data reporting

problems cited above for soy protein ingredients are still a problem in interpreting the soy food

literature.

Chien et al. conducted a critical study on the kinetics of genistein and its glucosides interconversions in a model system with kinetic estimates for this apparent first order reaction under dry

and moist heat treatments.38 Malonyl-genistin had the highest rate constant conversion to genistin

in both dry and moist systems; however, the magnitude of the rate constants were about ten times

faster in moist systems compared to dry. The rate constants for degradation in dry systems were:

MG to G > MG to AG >AG to G > MG to Gein ~ G to Gein ~ AG to Gein. The rate constants

for degradation in moist systems were: MG → G > MG → AG > AG → D2 > Gein → D4 > G →

D3 > AG → G >MG → D1 where Dx represents degradation products. The energies of activation

for moist heat conversions were in the following order: MG → G > G → D3 > Gein → D4 > MG

→ AG > AG → G > AG → D2. The energies of activation for dry heat conversions were in the

following order: G → D3 > MG → D1 > Gein → D4 > MG → AG >MG → G > AG → G. This

is the first report to give thermodynamic interpretation of isoflavone redistribution resulting from

thermal processing. These results mimic what has been observed in soy food systems for all 3

isoflavone forms.22

Setchell and Cole report isoflavone levels in 85 samples of soymilk and 2 types of soy protein

isolate produced commercially over a 3-year period.39 The data are not presented as individual

isoflavone forms but as total isoflavones or total malonyl, total β-glucoside, total acetylglucoside