Executive Summary

Since soybean isoflavones can be regarded as a functional food and their beneficial health effects are dosage dependent, it is important to control their levels in soybean seeds. Very different isoflavone levels and composition in soy products could be due to the genotype and growing environment of the soybeans used as starting material. Our results indicate that genetic background has greater effects on isoflavone levels and composition than does location. In green house experiments we found that environmental conditions such as temperature and soil moisture can greatly change the amount and composition of isoflavones in soybean seeds. Genotypes tested had different sensitivities to drought and temperature conditions as shown by alterations in seed isoflavone levels. The combination of drought and super-optimal temperature conditions resulted in a significant decrease in isoflavone accumulation and their composition in seeds of all genotypes tested. Genotype by environment interactions are significant for isoflavone concentrations but in our research were relatively small compared to genetic differences. We found that genetic diversity exists to either increase or decrease the current amount of isoflavones and to change the composition of isoflavones in U.S. cultivars. Our results should help to develop commercial lines for the farmer with very high and low isoflavone levels by either conventional breeding, marker assisted selection or biotechnological methods. These soybeans would be used by consumers that have a need for high or low levels of isoflavones, since high dietary isoflavones are beneficial for most but not all consumers as explained above. The high and low lines would also be ideal for clinical and laboratory research to determine which isoflavone or combination are most efficacious.

Research Progress

Soybean seeds contain several biologically active components (Eldridge and Kwolek, 1983; Messina and Messina, 1991) that make it an important functional food that can have positive effects on the health of humans and animals that consume it (Choi et al., 1996; Onozawa et al., 1999; Record et al., 1995; Wei et al., 1995). Even though soyfoods provide an array of important phytochemicals such as phenolic acids, protease inhibitors, saponins, and other compounds - most of the data indicate that many health promoting effects of soy foods are associated with consuming the components of soybean seeds called isoflavones, which are estrogen analogs (reviewed by Anderson and Garner 1997; Murkies et al. 1998; Setchell 1998; Anderson et al. 1999; Setchell and Cassidy 1999). In addition to their pro- or anti-estrogenic properties, isoflavones have important non-hormonal effects acting as tyrosine kinase inhibitors (Akiyama et al. 1987), natural anti-oxidants (Naim et al. 1976), anti-angiogenic agents (Fotsis et al. 1995), and by interacting with cell signal -transduction pathways (Bennetts et al. 1946). All these properties of isoflavones appear to cause their multiple beneficial effects on human health by decreasing the risk of certain diseases, including breast, prostate and colon cancer, osteoporosis, coronary heart disease and menopausal symptoms.

Isoflavones occur predominantly as glucosides or malonylglucosides and rarely as aglycones in soybean seeds (Ohta et al. 1979), and they are associated with the soybean protein fraction during seed processing. Ingested soybean isoflavones are hydrolyzed by intestinal glucosidases to release the aglycones daidzein, genistein and glycetein, and the extent of this conversion could be variable among individuals and with various diet components (Setchell et al. 1984; Setchell and Cassidy 1999).

Soybean products have long been a major part of the human diet in Asia, and the epidemiological data and animal and human observational studies largely support the hypothesis that soybean intake is associated with reduced risk for several types of cancers (Anderson et al. 1999; Messina et al. 1994; Kennedy 1995; Steele et al. 1995). Recently, the use of soy products has increased in the United States (where the average daily dietary intake of isoflavones is typically negligible) in order to prevent a range of hormone-relevant diseases.

High isoflavone intake is generally beneficial for most humans but the growth of certain breast cancers can be stimulated by isoflavones (Hsieh et al. 1998). The clinical and biological results of a isoflavone-rich diet seems to be dependent on the intake dosage and duration. The consumption of isoflavones from natural soy-foods might be more desirable than isoflavone supplements use, since supplements could be easy overdosed, which would result in their toxilogical levels in humans (Setchell et al. 1997).

Taking into account the important physiological effects of isoflavones and their dosage dependence, it is important to determine ways to control isoflavone levels in soybean seeds since these are important foods. Soy food produced from seeds with different levels and ratios of individual isoflavones could be of particular interest for various consumer groups.

Soy food products can have different isoflavone levels due to the variety of soybeans used as starting material, and the environment in which the soybeans were grown (Wang and Murphy, 1994 a, b).

Large differences in various isoflavone concentrations and composition in soybean seeds have been found in different varieties studied in the U.S. (Eldridge and Kwolek 1983; Wang and Murphy 1994 b; Hoeck et al. 2000), France (Lacombe et al. 1999 a, c), Korea (Choi et al. 1998) and Japan (Kitamura et al. 1991). However, only a small number of genotypes were tested in those studies.

Climatic conditions and agronomic practices might also be important factors causing the variation in isoflavone contents. The content and composition of isoflavones depends on the interaction of the genotype and sowing date (Aussenac et al. 1998). Isoflavone accumulation begins early in seed development and accelerates as the seed matures. A negative correlation between the isoflavone accumulation and temperatures during the seed filling period has been reported (Kitamura et al. 1991; Tsukamoto et al. 1995), which might at least in part explain why late maturing varieties have higher isoflavone levels than early maturing varieties. It has also been shown that a water deficit during the growing period can result in a significant decrease in both the isoflavone content and the weight of 1000 seeds as compared to the control (Lacombe et al. 1999b).

Several reports indicate the effect of various locations and years on isoflavone contents. Thus, total isoflavone levels varied in the range of 116-309 mg/g within four different cultivars grown in the same location, in the range of 46-195 mg/g in two of the same soybeans grown in four various locations and in the range of 245-363 mg/g in a single variety grown in the same location across four different years (1975, 1976, 1978 and 1979) (Eldridge and Kwolek 1983). Isoflavone contents in a single soybean cultivar ranged from 118 to 331 mg/g among three experimental years (1989-1991) and from 118 to 175 mg/g among various locations within the same 1991 year (Wang and Murphy 1994 b), indicating the greater impact of the crop year on the isoflavone contents compared to the location. When six soybean cultivars were grown at eight locations during two years, the performance of the two genotypes with the highest and lowest mean total isoflavone contents was relatively consistent among 16 environments, despite the significant interactions (Hoeck et al. 2000). Even though reported data indicate significant environmental effects on soybean seed isoflavone contents, more extensive evaluation of a larger number of genotypes across locations and years is apparently required.

The goal of our study is to determine the critical combination of genetic and environmental factors that will allow for the production of soybean seeds with uniformly high or low isoflavone levels since it appears that there could be a market for soybean with high or low isoflavone contents for different groups of consumers.

The objectives of our research were 1) to evaluate diversity in isoflavone concentration in selected U.S. cultivars and plant introductions from the USDA Soybean Germplasm Collection and 2) to study the environmental factors that can affect isoflavone concentration.

Methods

Twenty-five soybean seeds of each sample were ground for 5 min with a mechanical mill (Spex Industries, Inc., NJ, USA), and 200 mg of the powder was extracted with 1ml of 80% methanol at 20ºC with vigorous shaking (80rpm) in a G10 gyrotary shaker (New Brunswick Sci., NJ, USA for 12 h. The pellet obtained after centrifugation at 5,000 g for 10 min was extracted again with 1ml 80% methanol by shaking for 2 h and samples were centrifuged at 5,000 g for 10 min. The combined supernatants were centrifuged at 12,000 g for 10 min and then used for HPLC analysis.

Isoflavones were separated with a Waters HPLC system (Waters Corp., Milford, MA, USA) using a 53 mm x 7 mm, 3 m EPS C18 Alltech Rocket Column (Alltech Assoc., Deerfield, IL, USA) following a previously reported method (Graham 1991). A linear gradient composed of water (solvent A) (pH 2.8, adjusted with acetic acid) and acetonitrile (solvent B) was used. Following injection of 20 ml of sample, solvent B was increased from 5% to 12% over 7 min and then isocratic elution for 3 min occurred followed by an increase in solvent B to 35% over 15 min. The column was then washed with 95% acetonitrile for 3 min and equilibrated for 3 min at 5% B between runs. Total sample to sample time was 30 min. The solvent flow was 2.5 ml/min. A Waters 996 photodiode array detector was used to measure UV absorbance at 253 nm. Isoflavone standards (daidzein, genistein, glycitein and their glucosides) were purchased from LC Laboratories, Woburn, MA, USA. Malonyl forms were identified based on the response factors of the corresponding b-glucosides and appropriate correction for the molecular mass differences. Each sample was analyzed in 2 replicates. Standard errors between the same samples did not exceed 2%.

Results

Twelve publicly developed cultivars or ancestral lines of these cultivars were grown in 2 replications at three location in central Illinois for two years. Because it has been previously published that date of maturity can strongly effect isoflavone concentration, comparisons were made among entries with similar maturities. Five entries were in maturity group II and early maturity group III. The average dates of maturity for these lines range from September 11 to September 19. Seven entries were late maturity group III and early maturity group IV. The average dates of maturity for these lines ranges from September 24 to October 6.

Table 1. Entry means expressed as μg g-1 of dry weight for total and three major types of isoflavones average of three locations in 1999 and 2000.

Entry	Maturity	Total	Diadzein	Genistein	Glycitein
Peking	Oct. 6	4556	2359	1726	472
Williams 82	Sep. 25	3044	1317	1388	339
Williams	Sep. 24	2992	1294	1382	316
Pixie	Sep. 29	2761	1040	1448	273
Macon	Sep. 24	2719	968	1381	371
Cisne	Sep. 29	2190	860	1138	193
Rend	Sep. 29	1790	619	858	303
Dwight	Sep. 12	2858	1287	1254	317
Savoy	Sep. 11	2839	1150	1426	263
Wayne	Sep. 19	2103	880	943	280
Lincoln	Sep. 16	1988	739	950	300
Jack	Sep. 18	1549	516	749	284
LSD (0.05)	1.3	195	92	96	?22

From these data it is obvious that there are large genetic differences for isoflavone concentration and those differences are not always related to time of maturity. Glycitein is always the smallest component of isoflavone but there are apparent genetic differences that affect the relative concentrations of all three isoflavone types. The percentage of diadzein ranged from 52% in Peking to 33% in Jack. The percentage of genistein ranged from 52% Pixie to 38% in Peking. The lines with higher total concentration often have a higher percentage of diadzein but this is only a small sample and some variations to that trend exist here. Glycitein tends to be a higher percentage of the total when the total concentration is lower. Rend and Jack have the lowest total concentration of isoflavones and have 17 and 18% glycitein, respectively. Glycitein represented only 9% of the isoflavones in Cisne and Savoy. These data show that it should be possible to substantial change both the quantity and quality of isoflavones in soybean seeds and that differences are available within current commercial cultivars.

Previous reports have indicated that environmental influences on isoflavone concentrations are very large and some have suggested that these effects are much larger than genetic effects. Our data do not support such a conclusion. Table 2 shows the values for each of the six environments for the entries with the highest and lowest mean total concentrations. The total isoflavone concentration of Peking ranged from 3901 to 5949 μg g-1 of dry weight depending on the environment but in all environments Peking had the highest concentration. The isoflavone concentration of Jack ranged from 1275 to 1885 μg g-1 of dry weight. Jack had the lowest mean isoflavone concentration and was the lowest entry in 5 of the 6 environments. These data demonstrate that even when there are large variations in isoflavone concentrations caused by environmental conditions the relative rankings of entries can remain relatively constant.

Table 2. Entry means expressed as μg g-1 of dry weight and ranks (bold) by location and year for total and three major types of isoflavones in 1999 and 2000 for the entries with the highest and lowest overall total mean values.

Entry	Location	Year	Total	Diadzein	Genistein	Glycitein
Peking	Bellflower	1999	4003 1	2053 1	1693 4	258 6
Peking	Urbana	1999	4938 1	2524 1	2109 1	306 1
Peking	Hume	1999	3901 1	1942 1	1713 1	246 1
Peking	Bellflower	2000	4633 1	2451 1	1573 3	609 1
Peking	Urbana	2000	5949 1	3123 1	1966 1	861 1
Peking	Hume	2000	3914 1	2060 1	1302 4	553 1
Jack	Bellflower	1999	1695 12	583 12	847 12	266 4
Jack	Urbana	1999	1885 11	660 11	998 11	228 5
Jack	Hume	1999	1275 12	441 12	623 12	211 4
Jack	Bellflower	2000	1484 12	440 12	693 12	351 9
Jack	Urbana	2000	1173 12	376 12	523 12	274 12
Jack	Hume	2000	1782 12	598 12	812 12	372 6

The relative rankings of the concentrations of the three forms of isoflavones was also quite consistent for Peking and Jack. The greatest variation occurred in the ranking for the glycitein concentration for Jack even though the actual values changed very little. This variation is glycitein concentration was even more pronounced in the cultivar Pixie (Table 3). In most environments the total isoflavone concentration of Pixie was close to the mode with ranking between 6 and 8 out of 12 in four of the six environments. The rankings of diadzein and genistein were not quite as consistent but tended to be in the intermediate range. In 1999 the glycitein values were consistently among the lowest in the experiment but in 2000 there were much higher in absolute values and substantially higher in ranking. We do not have an explanation for this effect but these data indicate that conditions affecting glycitein may be different from those affecting the other isoflavone components. It is known that glycitein occurs primarily in the embryo and not in the cotyledons so that may be a factor.

Table 3. Entry means expressed as μg g-1 of dry weight and ranks (bold) by location and year for total and three major types of isoflavones in 1999 and 2000 for an entry with a highly variable glycitein content.

Entry	Location	Year	Total	Diadzein	Genistein	Glycitein
Pixie	Bell	1999	2926 6	1073 7	1704 2	150 11
Pixie	Urb	1999	2773 7	1024 6	1600 5	150 11
Pixie	Hume	1999	2123 6	786 7	1199 5	138 10
Pixie	Bell	2000	3460 2	1322 2	1692 1	447 3
Pixie	Urb	2000	2235 8	837 10	1067 7	331 8
Pixie	Hume	2000	3050 3	1199 3	1430 2	422 6

A selected set of cultivars and ancestral lines were grown in four environments in 1999 and 2000. The mean values are presented in Table 4. The entries are divided in six groups based on maturity with groups means also presented (Table 4). The group means demonstrate the effects of maturity especially for those very early maturing lines. The five lines that matured prior to August 7 averaged 758 μg g-1 whereas the 8 latest lines with an average maturity date of September 30 averaged 2437 μg g-1. By making comparisons within groups, these data also demonstrate the significant genetic component of isoflavone composition. Significant differences in total isoflavone concentration exist within each of the groups with the differences exceeding 200% in the set of latest maturing lines. Notable examples of lines that matured on the same date but had significant differences in isoflavone concentration are Mukden and Kanro, Lincoln and A.K. (Harrow), Macon and PI 88788, and Rend and Pixie. Only ancestral lines were included in the two sets of earliest maturing lines but in all of the four later maturing groups high and low concentrations of isoflavones were found in both the more recently developed cultivars and the ancestral lines. These data show that there is sufficient diversity in the ancestral lines to produce variation among our modern cultivars and that selection for high yield in commercial cultivars has not selectively shifted the isoflavone concentration.

Table 4. Entry means for total and three major types of isoflavones expressed as μg g-1 of dry weight averaged over four environments in 1999 and 2000.

Entry	Maturity	Total	Diadzein	Genistein	Glycitein
PI360955A	66	499	217	195	87
PI360955B	66	746	332	298	116
PI438477	68	787	238	347	202
Man. Brown	68	868	412	306	150
PI438471	69	888	384	366	138
Mean	68	758	317	302	139
Mandarin (O.)	84	806	326	353	126
Flambeau	76	814	332	333	150
Capital	87	1161	462	494	205
PI 180501	81	1363	528	603	232
Mean	82	1036	412	446	178
Korean	95	1066	405	491	170
Mukden	100	1399	547	596	256
Richland	98	1529	634	725	170
Dunfield	103	2379	955	1056	368
Kanro	100	2712	1383	1036	294
Bansei	101	2968	1500	1104	364
Savoy	102	2985	1222	1429	334
Dwight	102	3039	1381	1292	366
Mean	100	2259	1003	966	290
Jack	107	1581	518	757	306
Illini	106	2063	786	937	340
Lincoln	106	2053	764	957	331
Wayne	109	2202	911	970	322
A.K. (Harrow)	106	2237	854	1020	363
Mean	107	2027	767	928	333
Jogun	112	1800	851	778	171
Macon	115	2759	980	1345	434
Williams	115	2977	1282	1324	371
Williams 82	115	3002	1287	1323	392
PI 88788	115	4094	1773	1676	646
Mean	114	2926	1234	1289	402
Perry	121	1477	438	778	261
Rend	120	1839	637	874	327
PI 80837	123	1952	692	945	315
Cisne	120	2100	827	1053	220
FC 33243	119	2118	727	1009	382
PI 71506	128	2273	785	1177	311
Pixie	120	2879	1095	1447	337
Peking	124	4858	2539	1737	582
Mean	122	2437	968	1128	342
LSD (0.05)	1.4	200	94	90	31
Maturity is reported as days after May 31.

The consistency of ranking among the entries with the highest and lowest isoflavone concentrations was also observed with the lines tested in this experiment. Peking and PI 88788 had the two highest mean concentrations of isoflavones averaged over the four environments. The individual rankings showed the Peking had the highest concentration in each environment (Table 5) although the actual values ranged from 3914 to 5949 μg g-1. PI 88788 had the second highest concentration in three environments and was third in the fourth environment. The variation among environments in actual concentration was nearly 30% but as with Peking the relative values were very consistent. Similar consistency was found at the low end of the scale. Mukden and Korean had the two lowest isoflavone averages. Korean was also the lowest entry in each of the four environments (Table 5). Mukden was 28th of 31 in three environments and 30th in the fourth environment. These results show that environmental conditions do strongly influence the concentration of isoflavones. Even though the genotype by environment interaction was highly significant for all of the isoflavone measurements, the magnitude of those differences often has little impact on the relative rankings of the entries.

Table 5. Entry means expressed as μg g-1 of dry weight and ranks (bold) by location and year for total isoflavones in 1999 and 2000 for the entries with the two highest and two lowest entries total mean values.

Entry	Location	Year	Total	Rank
Peking	Urbana	1999	4938	1
Peking	Urbana	2000	5949	1
Peking	Bellflower	2000	4633	1
Peking	Hume	2000	3914	1
PI088788	Urbana	1999	4764	2
PI088788	Urbana	2000	4065	2
PI088788	Bellflower	2000	3817	2
PI088788	Hume	2000	3731	3
Mukden	Urbana	1999	1175	30
Mukden	Urbana	2000	1285	28
Mukden	Bellflower	2000	1614	28
Mukden	Hume	2000	1523	28
Korean	Urbana	1999	1122	31
Korean	Hume	2000	1270	31
Korean	Urbana	2000	955	31
Korean	Bellflower	2000	917	31

In 1999, 154 diverse germplasm accessions in maturity groups (MG) II to IV, mostly from China, were selected from the USDA Soybean Germplasm Collection for isoflavone analysis. These samples were in storage and had not all been grown in the same year. Based on these data, 36 accessions were identified to represent high and low amounts of total isoflavones, and total amounts of daidzein, genistein and glycitein (including the conjugate forms). These accessions plus 39 MG II to IV public soybean cultivars and major ancestral lines of northern U.S. cultivars were planted in two replications at three locations in 2000.

The relative values of the isoflavone concentrations of the exotic germplasm measured from a single sample taken from the USDA Soybean Germplasm Collection in 1999 compared to the replicated values from 3 locations in 2000 were very similar. These results were surprising since the samples analyzed in 1999 were unreplicated and the seeds were grown in variety of environments. The correlation between the actual values obtained in 1999 and the mean of six observations collected in 2000 was 0.86 for total isoflavones and the correlation between the rankings in the two years was 0.90 . The correlations between years for the individual isoflavone components were 0.85 for diadzein and genistein, and 0.80 for glycitein. Many of the actual values were quite different between years. PI 567387 had the highest value of the 154 samples in 1999 at 6114 μg g-1. In 2000, this accession had the third highest value but it was only 3962 μg g-1. PI 567305 , the accession with the highest value in 2000 at 4586 μg g-1 was fourth in 1999 with 5159 μg g-1. Selecting only the four highest samples based on 1999 data would have identified the three highest accessions in 2000. The top five in 2000 were among the top 12 in 1999. Selecting for low isoflavone content had similar results. The accession (PI 567162) with the lowest isoflavone concentration in 1999 (625 μg g-1) was also the lowest in 2000 (596 μg g-1). Selecting the 5 lowest accessions in 1999 would have identified the three lowest accessions in 2000 and the lowest five values in 2000 were among the eight lowest values in 1999. The accessions selected for high isoflavone values in 1999 all had lower values in 2000. The results were not as consistent for those selected for low concentration. For the accessions with the five lowest values in 2000, three values were lower in 2000 and two were higher. Regardless of these differences between years, it is very obvious from these data that there is strong genetic control of isoflavone concentration.

These results expanded the range of known genetic diversity for both commercial cultivars and exotic germplasm (Table 6). The maturity group II accessions had a range of maturity of 7 days and range of total isoflavone concentration of 596 μg g-1 (PI 567162) to 3665 μg g-1 (PI 567429A). The highest and lowest values were in exotic accessions that originated from China. The range in commercial cultivars was from Loda at 1406 μg g-1 to Dwight at 2956 μg g-1 even though these cultivars differed in maturity by only 1 day (Table 6). The lowest and highest values in maturity group III entries were also among the exotic accessions with PI 567367 at 894 μg g-1 and PI 567491A at 4199 μg g-1 . The range was slightly larger among the group III accessions than among group II accessions and the extreme values were also slightly higher. The extremes among the group III cultivars were Jack (1480 μg g-1) and Williams 82 (2842 μg g-1), which were also identified as extremes in the previous study. The greatest range and highest value were found in maturity group IV entries. PI 567366A was the lowest value at 914 μg g-1 whereas Peking, the highest entry (4832 μg g-1) among ancestral lines and cultivars in the previous study, was still the highest entry when we included selected exotic accessions. Among the cultivars tested Perry had the lowest concentration at 1422 μg g-1. It was also identified as the lowest concentration among the ancestral lines of northern cultivars. Ripley had the highest concentration among the cultivars at 3543 μg g-1 (Table 6).

Table 6. Entry means expressed as μg g-1 of dry weight for total isoflavones and the three major classes of isoflavones averaged over three locations in 2000 and the rank of each value.

Line		Maturity	Rank	Total	Rank	Daidzein	Rank	Genistein	Rank	Glycitein	Rank
Flambeau	0	73	76	892	75	372	73	348	73	172	70
Manitoba Brown	0	66	77	894	74	431	64	301	76	161	72
Mandarin (O.)	1	82	75	787	76	334	75	310	75	144	75
Capital	1	85	74	1124	68	473	61	440	68	211	67
PI567162	2	94	73	596	77	278	77	201	77	118	77
PI567161	2	100	61	894	72	410	68	331	74	154	74
PI567323A	2	99	62	1296	62	574	54	502	65	221	65
PI567166	2	96	69	1321	60	478	60	547	61	295	46
Loda	2	99	64	1406	58	514	57	640	55	252	58
Conrad	2	98	65	1409	57	556	55	619	56	234	62
Richland	2	97	68	1509	54	640	51	665	54	205	68
PI567313	2	97	67	1528	53	698	48	581	58	248	60
PI567481	2	96	71	1532	52	555	56	710	50	267	55
CN290	2	100	59	1619	50	574	53	710	51	335	38
IA2038	2	100	58	1803	46	831	43	717	48	255	57
Hoyt	2	100	60	1930	43	832	42	788	42	310	45
Century	2	101	56	2280	35	1004	29	950	35	326	43
Elgin	2	95	72	2328	34	1072	25	925	38	332	42
Preston	2	98	66	2349	32	997	30	955	34	398	27
Hack	2	96	70	2513	28	1087	23	1073	29	353	37
Savoy	2	101	55	2694	24	1094	21	1207	22	393	28
Burlison	2	101	54	2708	22	1126	19	1214	19	368	32
Dwight	2	100	57	2956	16	1348	13	1191	23	417	25
PI567429A	2	99	63	3665	7	1839	5	1291	15	536	6
PI567367	3	112	36	894	73	298	76	379	71	218	66
PI567360	3	114	31	941	70	359	74	415	69	167	71
PI567452	3	109	41	1127	67	385	69	466	67	276	52
PI567524	3	109	42	1158	66	378	70	551	60	229	63
PI567453	3	109	40	1176	65	372	72	519	63	286	47
PI567511	3	108	44	1256	63	461	63	514	64	281	51
PI567168	3	104	52	1353	59	680	49	493	66	180	69
Jack	3	103	53	1480	55	471	62	676	53	332	41
Logan	3	106	48	1713	49	720	47	719	47	274	53
PI567543A	3	107	46	1733	48	626	52	784	43	323	44
PI567372A	3	110	38	1792	47	771	45	739	46	282	50
IA3010	3	108	45	1896	45	860	39	766	45	270	54
PI567510A	3	113	32	1904	44	988	31	680	52	236	61
Lincoln	3	104	51	2011	40	744	46	906	39	361	34
Zane	3	105	49	2147	38	931	36	856	41	360	35
Wayne	3	107	47	2224	36	915	37	940	36	370	31
PI567516A	3	115	28	2602	26	936	35	1213	20	453	16
Sprite	3	109	43	2694	25	1029	28	1243	18	422	22
Macon	3	113	33	2742	21	972	33	1275	16	495	9
Williams	3	113	34	2806	19	1199	16	1184	24	423	21
Williams 82	3	114	30	2842	18	1193	17	1207	21	443	18
PI567286	3	112	35	2983	15	1564	9	1001	32	418	24
PI567519	3	115	29	3013	14	1296	15	1257	17	460	14
PI567461	3	105	50	3756	6	1944	3	1343	11	468	13
PI567465	3	110	39	3865	5	2063	2	1362	9	440	19
PI567491A	3	111	37	4199	3	1928	4	1485	5	787	2
PI567366A	4	118	25	914	71	373	71	403	70	138	76
PI567359	4	118	24	958	69	424	66	374	72	160	73
PI567514A	4	120	19	1215	64	421	67	544	62	250	59
PI567354	4	127	3	1297	61	502	59	574	59	221	64
Perry	4	120	21	1422	56	426	65	712	49	285	48
PI567289A	4	124	8	1615	51	512	58	769	44	333	40
Rend	4	121	14	1941	42	676	50	903	40	362	33
PI567374	4	128	1	1997	41	847	40	603	57	547	5
Cisne	4	120	22	2043	39	813	44	968	33	263	56
PI567279A	4	120	20	2171	37	959	34	930	37	282	49
Custer	4	121	17	2339	33	838	41	1074	28	428	20
PI567529	4	121	16	2397	31	873	38	1106	27	419	23
Ina	4	121	11	2412	30	1039	27	1040	30	333	39
HS93-4118	4	121	13	2488	29	1093	22	1036	31	359	36
Clark	4	116	27	2586	27	1084	24	1123	26	380	30
Lawrence	4	117	26	2706	23	1160	18	1160	25	386	29
PI567287	4	121	12	2781	20	981	32	1350	10	450	17
Pixie	4	121	18	2915	17	1119	20	1396	8	400	26
Franklin	4	119	23	3113	13	1063	26	1538	4	512	7
PI567476	4	122	10	3129	12	1315	14	1308	14	506	8
PI567311A	4	127	2	3201	11	1385	12	1329	13	487	11
PI567331	4	125	6	3362	10	1404	11	1467	7	491	10
PI567294	4	126	4	3440	9	1622	6	1335	12	484	12
Ripley	4	123	9	3543	8	1606	7	1483	6	455	15
PI567387	4	125	7	3962	4	1569	8	1660	2	733	3
PI567305	4	121	15	4586	2	1450	10	1665	1	1470	1
Peking	4	125	5	4832	1	2544	1	1614	3	674	4
Lsd (0.05)		2.4		311		121		99		210

In maturity groups II through IV, which were represented by both cultivars and exotic germplasm, we identified exotic accessions that were significantly higher or lower than the extreme values identified in the cultivars. Peking, the highest entry in maturity group IV, is an ancestral line from China but based on pedigree analysis has contributed less than 1% of the genes to current cultivars. The accessions from China, including Peking, identified in this research could be used to increase the isoflavone concentration of U.S. cultivars. Perhaps more surprising than the accessions with concentrations much higher than the cultivars were the accessions with very low concentrations. The accessions with the lowest concentrations in each maturity group were not significantly different from each other. To identify accessions in maturity group III and IV that have isoflavone concentrations as low as accessions in maturity group 00 that matured more than 50 days earlier is a novel and significant discovery that would make it possible to significantly lower the isoflavone concentration in commercial cultivars.

These data also indicate that there is significant genetic diversity for the quality of isoflavones among commercial cultivars and further genetic changes should be possible. The range in percentages of each of the three major isoflavone types within the cultivars ranged from 31 to 45% for diadizein, 38 to 47% for genistein, and 15 to 25% for glycitein. Differences were found within the exotic germplasm with percentages as high as 52% for diadzein and 34% for glycitein and as low as 26% for genistein.

Five cultivars were selected for a greenhouse study to quantify the effects of soil moisture and air temperature during the final weeks of seed filling. These cultivars were selected based on previous data that indicated these lines possessed a range of isoflavone concentration as well as a diverse responses to environmental changes. These entries were also similar in maturity with all being in maturity group II. Queen and Imari are commercial cultivars grown in France and the other three cultivars were developed at the University of Illinois. All entries were planted in five replications with a single plant in a _11_ in. pot as an experimental unit. All plants were grown under well watered conditions and 18/28º C day/night temperatures until approximately the R6 growth stage. At that time the plants were moved into three temperature regimes: 13/23º C, 18/28º C or 28/38º C. Half of the plants in each treatment continued to receive optimum water and half were maintained at approximately 30% of well-watered conditions. The experiment was repeated twice.

Averaged across all environments and experiments there were significant differences among cultivars for all traits measured (Table 7). Under these conditions Jack and Dwight were much later in maturity than the other three entries and were the highest yielding cultivars. Queen and Jack were much taller than the other entries at 160 and 143 cm, respectively. Queen and Dwight had the highest and similar total isoflavone concentrations. Queen had significantly more diadzein and glycitein than Dwight but Dwight had a higher concentration of genistein. Jack and Loda were lower in total isoflavone and genistein concentration than the highest two cultivars but not significantly different from each other. Loda had a significantly higher daidzein concentration whereas Jack had a significantly higher glycitein concentration.

Table 7. Isoflavone concentration of entry means expressed as μg g-1 of dry weight for total isoflavones and the three major classes of isoflavones and other plant data averaged over 5 replications per experiment, two experiments, three air temperature regimes and two soil moisture treatments.

Entry	Maturity date (days)	Plant height (cm)	Seed yield (gm/plt)	100 seed weight	Diadzein	Genistein	Glycitein	Total
Dwight	58	92	44	12.9	1272	1441	268	2981
Imari	49	89	28	18.6	527	652	113	1293
Jack	59	143	42	12.4	549	1031	276	1856
Loda	44	101	35	19.0	733	1089	216	2038
Queen	45	161	36	16.5	1391	1293	333	3017
Lsd (0.05)	2	10	4	1.6	111	107	27	211

The plants were grown under the same conditions for most of the experiment in an attempt to reduce the effects of the environmental treatments on overall plant growth and productivity; however, the treatments did effect many plant traits (Table 8). There was slight but significant hastening of maturity under low temperatures and an opposite effect with high temperatures. Dry soil conditions significantly hastened maturity. Temperature significantly affected plant height. The intermediate temperature produced the tallest plants and the low temperature the shortest plants. Seed yield has slightly but significantly higher with low temperatures or high soil moisture. One hundred seed weight was not affected by either the temperature or soil moisture treatment. Total isoflavone concentration was significantly changed by all main treatment effects with isoflavones increasing with decreasing temperature and increased soil moisture.

Table 8. Isoflavone concentration means for temperature and soil moisture treatments expressed as μg g-1 of dry weight for total isoflavones and the three major classes of isoflavones averaged over 5 cultivars, 5 replications per experiment, and two experiments.

Treatment	Maturity date (days)	Plant height (cm)	Seed yield (gm/plt)	100 seed weight	Diadzein	Genistein	Glycitein	Total
Low temp	49	108	40	16.1	1425	1686	260	3371
Inter. temp	51	129	36	15.7	783	967	243	1994
High temp	53	115	35	15.9	475	950	221	1346
Lsd (0.05)	2	8	3	1.2	88	84	21	166
Low moist.	49	115	35	15.4	830	1004	234	2405
High moist.	54	119	40	16.3	958	1198	248	2069
Lsd (0.05)	1	6	3	1.0	71	69	17	135

Daidzein showed the same pattern as total isoflavones for both changes in temperature and soil moisture. Genistein was significantly higher with higher soil moisture but only with the lowest temperature was it significantly different from the other two temperature treatments. Glycitein was the most stable with no significant change due to soil moisture and a slight, but significant reduction with the highest temperature.

High temperature and wet soils delayed maturity the most and low temperatures and dry soil had the greatest effect in hastening maturity (Table 9). Most of the treatment combination did not significantly affect yield. Plant grown under high temperature and low soil moisture were significantly lower yielding than all other treatments. As with main effects, no treatment combinations significantly affected 100 seed weight. The highest isoflavone concentration was produced with low temperatures and high soil moisture. The same conditions that one would anticipate favoring yield under field conditions and which did produce the highest yield in these greenhouse experiments. The opposite treatment, high temperatures and low soil moisture, produced the lowest total isoflavone concentration and the lowest yield. As would be predicted from the main effects presented in Table 8 soil moisture was most important in determining isoflavone concentration and within temperature treatments high soil moisture always produced higher concentrations that low soil moisture. All six treatments were significantly different from each other except for the two intermediate temperature treatments. Diadzein and genistein followed the same pattern as total isoflavones. Glycitein was less affected by the changes and it is difficult to identify conditions that would consistently increase glycitein, although high temperature is more likely to lower the concentration.

Table 9. Isoflavone concentration means for the six treatments expressed as μg g-1 of dry weight for total isoflavones and the three major classes of isoflavones averaged over 5 cultivars, 5 replications per experiment, and two experiments.

Treatment	Maturity date (days)	Plant height (cm)	Seed yield (gm/plt)	100 seed weight	Diadzein	Genistein	Glycitein	Total
HD	50	117	31	15.3	388	491	209	1089
HN	56	113	39	16.6	561	810	232	1604
LD	47	104	37	15.7	1315	1569	263	3147
LN	52	111	43	16.4	1535	1802	258	3595
ND	49	124	35	15.3	788	953	231	1971
NN	52	134	37	16.1	779	982	255	2016
lsd 0.05	2	11	5	1.7	122	117	29	231
HD = High temperatures, dry soil HN = High temperatures, wet soil LD = Low temperatures, dry soil LN = Low temperatures, wet soil ND = Intermediate temperatures, dry soil NN = Intermediate temperatures, wet soil

Examing the effects of soil moisture on individual cultivars reveals a cultivar by soil moisture interaction (Table 10). None of the U.S. cultivars had a significant change in isoflavone component due to changes in soil moisture and most of the values were nearly identical across treatments. Imari had a significant reduction in genistein concentration and a sizeable but non-significant reduction in total isoflavones. Queen had large and significant reductions in daidzein, glycitein and total isoflavone concentrations and a nearly significant reduction in genistein. All of the cultivars matured significantly earlier under reduced soil moisture conditions, but none had significant changes in plant height. Imari and Queen both had significant reductions in seed yield as did Jack. Surprisingly there were no significant changes in 100 seed weight but Queen had the largest reduction.

Table 10. Isoflavone concentration means for the entry by soil moisture treatments expressed as μg g-1 of dry weight for total isoflavones and the three major classes of isoflavones averaged over 5 replications per experiment, two experiments, and three temperature regimes.

Entry	Soil	Maturity date (days)	Plant height (cm)	Seed yield (gm/plt)	100 seed weight	Diadzein	Genistein	Glycitein	Total
Dwight	D	56	92	45	12.5	1306	1381	253	2940
Dwight	N	60	92	43	13.4	1238	1501	282	3021
Imari	D	46	87	23	18.5	412	495	106	1013
Imari	N	51	90	34	18.8	643	809	120	1572
Jack	D	56	139	39	11.9	584	1033	276	1892
Jack	N	62	148	46	12.9	515	1029	277	1821
Loda	D	42	99	34	18.9	798	976	227	2001
Loda	N	47	103	37	19.1	668	1202	205	2075
Queen	D	42	159	32	15.4	1052	1137	310	2499
Queen	N	48	163	39	17.5	1729	1449	357	3535
lsd 0.05		3.2	15.6	6.6	2.3	345.2	330.0	39.2	662.0
D = dry soil conditions N = wet soil conditions

The effects of temperature are more uniformly expressed across all cultivars but some differences do exist (Table 11). Differences among all temperature treatments for total isoflavones were significant except for the differences between the intermediate and high temperatures for Imari and Loda. Loda and Dwight had the smallest percentage increases between the high and intermediate temperatures. The changes between the high and low temperatures were most dramatic ranging from 110% from Dwight to 218% for Imari. Lower temperatures have a consistent and large influence in raising total isoflavone concentrations. The effect of daidzein was not as consistent between the intermediate and high temperatures. That difference was not significant for Imari, Jack, or Loda. All of the difference between the intermediate and low temperatures were significant with the largest absolute value (904 μg g-1) and percentage increase (276%) obtained in the cultivar Queen. All of the temperature treatments produced significant changes in genistein concentration with the greatest absolute change (895 μg g-1) and percentage increase (269%) occurring in the cultivar Jack. Temperature did not significantly affect the time of maturity for Dwight, Jack and Loda, but higher temperatures did delay the maturity of Imari and Queen. The plant height of Dwight and Imari, the two shortest cultivars, were not changed by temperature but the other cultivars had significantly taller plants with intermediate temperatures. Only Imari and Jack had seed yield significantly affected by temperature but Imari yielded the most with intermediate temperatures and Jack yielded the least with intermediate temperatures.

Table 11. Isoflavone concentration means for the entry by air temperature treatments expressed as μg g-1 of dry weight for total isoflavones and the three major classes of isoflavones averaged over 5 replications per experiment, two experiments, and two soil moisture regimes.

Entry	Temp	Maturity date (days)	Plant height (cm)	Seed yield (gm/plt)	100 seed weight	Diadzein	Genistein	Glycitein	Total
Dwight	H	60	87	41	12.4	815	983	240	2039
Dwight	N	57	100	44	12.7	1101	1245	270	2615
Dwight	L	57	89	46	13.7	1900	2096	292	4288
Imari	H	53	91	23	20.8	278	296	95	669
Imari	N	46	91	34	17.9	455	520	111	1085
Imari	L	47	85	28	17.2	850	1141	134	2124
Jack	H	59	147	42	12.6	249	476	247	971
Jack	N	58	165	35	11.6	457	861	272	1590
Jack	L	60	118	50	13.1	942	1756	310	3008
Loda	H	45	97	34	17.5	436	731	211	1378
Loda	N	46	117	32	19.7	569	964	211	1744
Loda	L	42	90	39	19.7	1193	1572	227	2992
Queen	H	48	153	36	16.5	596	768	310	1675
Queen	N	46	172	34	16.4	1336	1246	351	2933
Queen	L	42	157	37	16.5	2240	1864	339	4443
lsd 0.05		3.9	17.9	7.9	2.7	283.5	213.7	1.0	469.3
H = High temperatures L = Low temperatures N = Intermediate temperatures

Table 12. Isoflavone concentration means for the entry by treatments expressed as μg g-1 of dry weight for total isoflavones and the three major classes of isoflavones averaged over 5 replications per experiment, and two experiments.

Entry	Temp	Maturity date (days)	Plant height (cm)	Seed yield (gm/plt)	100 seed weight	Diadzein	Genistein	Glycitein	Total
Dwight	LN	60	91	48	14.3	1769.2	2109.4	263.4	4142
Dwight	LD	53	87	45	13.2	2031.1	2081.7	321.3	4434
Dwight	NN	56	106	37	12.0	918.4	1088.6	298	2305
Dwight	ND	59	94	51	13.4	1283	1401.2	241.6	2926
Dwight	HN	64	80	45	14.0	1027.3	1304.9	284.5	2617
Dwight	HD	56	95	38	10.8	603	661.2	196.3	1461
Imari	LN	51	80	31	17.5	1031.4	1345.2	132.9	2510
Imari	LD	43	89	25	17.0	668	936.1	134.7	1739
Imari	NN	47	100	42	20.2	534.7	658.4	118.5	1312
Imari	ND	45	82	25	15.6	374.5	382.1	102.9	859
Imari	HN	55	91	27	18.7	361.8	424.3	108.9	895
Imari	HD	51	91	20	22.8	193.4	166.9	81.7	442
Jack	LN	62	123	53	13.6	819	1707.3	298.5	2825
Jack	LD	57	113	48	12.7	1064.6	1804.8	321.7	3191
Jack	NN	61	171	35	11.5	440.9	799.4	278	1518
Jack	ND	55	160	34	11.7	472.9	923.3	266	1662
Jack	HN	61	151	49	13.7	283.8	580.6	254.1	1119
Jack	HD	57	143	35	11.5	214.1	370.7	239	824
Loda	LN	44	97	42	19.9	997.2	1732.8	213.1	2943
Loda	LD	41	83	36	19.6	1389.6	1411.8	239.9	3041
Loda	NN	48	118	33	19.0	577.6	1037.9	210.7	1826
Loda	ND	43	115	32	20.5	561.1	890.1	210.6	1662
Loda	HN	48	95	35	18.5	428.5	834.8	192.1	1455
Loda	HD	42	100	33	16.5	442.7	626.8	230.2	1300
Queen	LN	44	164	41	16.9	3057.6	2116.3	380.9	5555
Queen	LD	40	150	34	16.0	1421.8	1611.9	296.7	3330
Queen	NN	49	174	37	17.6	1424.3	1325.2	368.9	3118
Queen	ND	43	171	32	15.2	1246.7	1166.8	333.8	2747
Queen	HN	52	150	40	18.1	704	905.3	321	1932
Queen	HD	44	156	31	14.9	488.8	630.8	298.6	1418
Lsd (0 .05)	4.6	26	11	4	283	273	62	534

Queen produced the most consistent response to changes in temperature and soil moisture with a consistent increase with each incremental change to lower temperature or increase soil moisture. It changed from 1418 μg g-1 of total isoflavones under hot and dry conditions to 5555 μg g-1 under cool, moist conditions. Queen produced the highest level of isoflavones of any entry. Imari also produced the highest isoflavone concentration with the lowest temperature and highest soil moisture. Total isoflavones in both Imari and Queen had a significant increase by increasing soil moisture under the lowest temperature regime. The remaining three cultivars had small non-significant decreases in total isoflavone concentrations with this treatment change. All cultivars had the lowest isoflavone concentrations with high temperatures and low soil moisture. Imari was the most sensitive to environmental changes with nearly a 470% change from the lowest to the highest values. It was also significantly lower in isoflavones than the other entries at most treatment levels. The smallest percentage change (134%) induced by environmental conditions occurred with Loda.

Imari and Queen produced the highest levels of diadzein and genistein at the low temperature, high moisture treatment. Loda, Jack and Dwight all had reductions in diadzein concentration when the soil moisture was increases at the lowest temperature regime. This decrease was actually significant for Loda. All entries except Jack produced the most genistein at the low temperature, high moisture treatment and the increases between the low and high soil moisture were statistically significant for Imari, Loda, and Queen. Percentage and absolute value changes were smallest with glycitein. Significant differences in glycitein concentration with changes in environmental conditions were found for all cultivars except Loda.

Conclusions

These results represent the most thorough examination of the genetic and environmental influences on isoflavones in soybean seed that has been conducted. We have identified large genetic differences in total isoflavones and in the various components of seed isoflavones. There is large variation within the current commercial cultivars that could be exploited but the diversity within exotic germplasm is much greater. It would be possible to significantly raise or lower the isoflavone level in future cultivars. Temperature during seed filling is a major environmental factor influencing isoflavone concentration. In general lower temperatures increase isoflavone levels but the magnitude of response seems to be cultivar dependent. Soil moisture can also change isoflavone levels with moisture stress reducing isoflavone concentrations. This response is very highly dependent on the cultivar. Most of the cultivars that we tested did not show a significant response to reduced soil moisture. Even though environmental conditions that we produced in the greenhouse could change the total isoflavone concentration by over 400%, the ranking of cultivars by average isoflavone concentration was consistent. Temperature and soil moisture can change the level of isoflavones in the seeds but the potential for isoflavone production is strongly controlled by genetics.

Publications

Nelson, R. L., V. Lozovaya, A. Lygin, and J. Widholm. 2001. Variation in Isoflavones in Seeds of Domestic and Exotic Soybean Germplasm. In 2001 Agronomy Abstracts. ASA, Madison, WI

Nelson, R.L., A.V. Lygin, V.V. Lozovaya, A.V. Ulanov, and J.M. Widholm. 2002. Genetic and environmental control of soybean seed isoflavone levels and composition. p. p511. Proc. 9th Biennial Conference of the Cellular and Molecular Biology of the Soybean. 11-14 Aug. 2002. Urbana, IL.

V. Lozovaya, A. Lygin, J. Widholm and R. Nelson. 2002. Effects of environmental factors on isoflavone amount and composition in soybean. In 2002 Agronomy Abstracts. ASA, Madison, WI. seeds.

Presently we are finalizing the preparation of three manuscripts for publication based on results obtained in this study.

References

Akiyama T., Ishida J., Odawaga H., Watanabe S., Itoh N., Shibuya M and Fukumi Y. (1987) Genistein, a specific inhibitor of tyrosine-specific protein kinases. J. Biol. Chem. 262: 5592-5595.

Anderson JJB, Anthony M, Messina M, and Garner SC. Effects of phyto-oestrogens on tissues. Nutr Research Reviews 12, 75-116. 1999.

Anderson J.J.B. and Garner S.C. (1997) Phytoestrogens and human function. Nutr. Today 32:232-239.

Aussenac T. Lacombe S. and Dayde J. (1998) Quantification of isoflavones by capillary zone electrophoresis in soybean seeds: effects of variety and environment. Am. J. Clin.Nutr. 68(suppl.):1480S-5S.

Bennets H.W., Underwood E.J. and Shier F.L. (1946) A specific breeding problem of sheep on subterranean clover pastures in Western Australia. Aust. Vet. J. 22:2-12.

Choi J.-S., Kwon T.-W., and Kim J.-S. (1996) Isoflavone contents in some varieties of soybean. Foods and Biotechnology, 5:167-169.

Eldridge A.C. and Kwolek W.F. (1983) Soybean isoflavones: effect of environment and variety on composition. J. Agric. Food Chem. 31: 394-396.

Fotsis T, Pepper M, Adlercreutz H, Hase T, Montesano R, and Schweigerer L. Genistein, a dietary ingested isoflavonoid, inhibits cell proliferation and in vitro angiogenesis. J Nutr 125, 790S-797S. 1995

Graham, T.L. (1991) A rapid, high resolution high performance liquid chromatography profiling procedure for plant and microbial aromatic secondary metabolites. Plant Physiol. 95:584-593.

Hoeck J.A., Fehr W.R., Murphy P.A. and Welke G.A. (2000) Influence of genotype and environment on isoflavone contents of soybean. Crop Sci. 40:48-51.

Hsieh C-Y., Santell R.C., Haslam S.Z., Helferich W.G. (1998) Estrogen effects of genistein on the growth of estrogen receptor-positive human breast cancer (MCF-7) cells in vitro and in vivo. Cancer research 58:3833.

Kennedy AR. The evidence for soybean products as cancer preventive agents. J Nutr 125, 733S-743S. 1995.

Kitamura K., Igita K., Kikuchi A., Kudoi S. and Okubo K. (1991) Low isoflavone content in some early maturing cultivars, co-called "summer-type soybeans" (Glycine max (L) Merrill) Japan J. Breed 41:651.

Lacombe S., Aussenac T., La Droitte Ph. and Dayde J. (1999a) Variability of isoflavone content in 22 European cultivars and effects of transformation process on isoflavones content in traditional soyfoods. Abstr. of WSRCVI, Aug. Chicago

Lacombe S., Aussenac T., La Droitte Ph. and Dayde J. (1999b) Effects of environmental factors on isoflavone content and composition in soybean seeds. Abstr. of WSRCVI, Aug. Chicago, 709.

Lacombe S., Aussenac T., La Droitte Ph. and Dayde J. (1999c) Isoflavones, "health" compounds in soybean seeds and soyfoods: variation of content and composition in raw material and soy-products. Abstr. of WSRCVI, Aug. Chicago, 696.

Messina M. and Messina V. (1991) Increasing use of soybean of soyfoods and their potential role in cancer prevention. J. Am. Diet. Assoc. 91: 836.

Messina MJ, Persky V, Setchell KDR, and Barnes S. Soy intake and cancer risk: a review of the in vitro and in vivo data. Nutr Cancer 21, 113-131. 1994.

Murkies A.L.,Wilcox G. and Davis S. (1998) Phytoestrogens. J. Clin. Endocrinol. Metab. 83:297-303.

Naim M., Gestetner B., Bondi A.,, Birk Y (1976) Antioxidative and antihemolytic activity of soybean isoflavones. J. Agric. Food Chem. 22:806.

Ohta N., Kuwata G., Akahori H., and Watanabe T. (1979) Isoflavonoid constituents of soybeans and isolation of a new acetyldaidzin. Agric. Biol. Chem. 43:1415-1419.

Onozava M., Kawamori T., Baba M., Fukuda K., Toda T., Sato H., Ohtani M., Akaza H., Sugimura T. and Wakabayashi K. (1999). Effects of a soybean isoflavone mixture on carcinogenesis in prostate and seminal vesicles of F344 rats, Jpn. J. Cancer Res. 90:393-398.

Record L.R., Dreozti L.E. and McInerney J.K. (1995) The antioxidant activity of genistein in vitro. J. Nutr. Biochem. 6:481.

Setchell K.D.R. (1998) Phytoestrogens:biochemistry, physiology and implications for human health of soy isoflavones. Am. J. Clin. Nutr. 68:1333S-1346S.

Setchell K.D.R. and Cassidy A. (1999) Dietary isoflavones: biological effects and relevance to human health. J. Nutr. 129:758-767.

Setchell K.D.R., Borriello S.P., Hulme P., Kirk D.N. and Axelson M. (1984) Non-steroidal oestrogens of dietary origin:possible roles in hormone dependent diseases. Am. J. Clin. Nutr. 40:569-578.

Steele VE, Pereira MA, Sigman CC, and Kelloff GJ. Cancer chemoprevention agent development strategies for genistein. J Nutr 125, 713S-716S. 1995.

Tsukamoto C., Shimada S., Igita K. et al. (1995) Factors affecting isoflavone content in soybean seeds: changes in isoflavones, saponins, and composition of fatty acids at different temperatures during seed development. J. Agric. Food Chem. 43:1184-1192.

Wang H.-J. and Murphy P.A. (1994a) Isoflavone content in commercial soybean foods. J. Agric. Food Chem. 42: 1666-1673.

Wang H.-J. and Murphy P.A. (1994b) Isoflavone composition of American and Japanese soybeans in Iowa: Effects of variety, crop year, and location. I. Agric. Food Chem. 42:1674-1677.

Wei H., Bowen R., Cai R., Barnes S. and Wang Y. (1995) Antioxidant and antipromotional effects of the soybean isoflavone genistein. Proc.Soc. Exptl. Biol. Med. 208:124