Crop Science Illumina
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 QUICK SEARCH:   [advanced]


     


Published online 23 September 2005
Published in Crop Sci 45:2203-2211 (2005)
© 2005 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Primomo, V. S.
Right arrow Articles by Rajcan, I.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Primomo, V. S.
Right arrow Articles by Rajcan, I.
Agricola
Right arrow Articles by Primomo, V. S.
Right arrow Articles by Rajcan, I.
Related Collections
Right arrow Soybean
Right arrow Plant and Environment Interactions
Right arrow Crop Genetics

CROP BREEDING, GENETICS & CYTOLOGY

Agronomic Performance of Recombinant Inbred Line Populations Segregating for Isoflavone Content in Soybean Seeds

Valerio S. Primomoa, Vaino Poysab, Gary R. Ablettc, Chung-Ja Jacksond and Istvan Rajcana,*

a Dep. of Plant Agriculture, Crop Science Division, Univ. of Guelph, Guelph, ON, Canada N1G 2W1
b Agric. and Agri-Food Canada, 2585 Highway 20 East, Harrow, ON, Canada N0R 1G0
c Ridgetown College, Univ. of Guelph, Ridgetown, ON, Canada N0P 2C0
d Guelph Center for Functional Foods, Lab. Services Division, Univ. of Guelph, 95 Stone Road West, Guelph, ON, Canada N1H 8J7

* Corresponding author (irajcan{at}uoguelph.ca)


    ABSTRACT
 TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Soybean [Glycine max (L.) Merrill] seeds contain isoflavones, which have been associated with positive health effects in human adults but a negative effect on infants. Increasing or decreasing isoflavone content in the seed would be desirable; however, it is not known what impact this would have on agronomic and other seed quality traits. The main objective of this study was to determine if isoflavone content in soybean seeds was associated with changes in agronomic and seed quality traits. Recombinant inbred lines (RILs) with "high" (n = 35), "intermediate" (n = 35), and "low" (n = 35) isoflavone content, were selected from three F4:5 populations grown in 2000 and planted at three locations in southern Ontario in 2002. There were significant differences among populations, environments, and their interaction for isoflavone content. "High" and "low" phenotypic classes were significantly different for maturity in all three populations. Despite a positive correlation between yield and isoflavones, several RILs with decreased isoflavone content (970 µg g–1) were identified with yields (4222 kg ha–1) comparable to high-yielding cultivars. Isoflavone content had minimal effects on oil content, seed quality, and weight. Population 1 showed a significant negative association between isoflavone and protein content, whereas Pop. 2 and Pop. 3 did not. RILs were identified with high isoflavone and protein content ranging from 1746 to 1851 µg g–1 and 431 to 442 g kg–1, respectively. It was possible to develop soybean with desirable isoflavone content in the seed and superior agronomic and seed quality traits.

Abbreviations: HPLC, high performance liquid chromatography • NIR, near infra-red reflectance • PDA, photodiode array • RIL, recombinant inbred line • SAS, statistical analysis software


    INTRODUCTION
 TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
ISOFLAVONES are secondary metabolites synthesized by 2-hydroxyisoflavanone synthase (IFS) and are found almost exclusively in legumes (Heller and Forkmann, 1994). Two IFS genes (IFS1 and IFS2) encoding functional proteins have been identified and characterized from soybean (Jung et al., 2000). Studies have shown that IFS1 and IFS2 are expressed primarily in soybean roots and seeds, respectively (Dhaubhadel et al., 2003; Subramanian et al., 2004). Soybean seeds contain three major isoflavone aglycones (i.e., daidzein, genistein, and glycitein) and three each of their corresponding glucosides, acetyl-, and malonyl-ester glucosides (Jackson et al., 2002). In soybean, isoflavones have been shown to inhibit pathogen attack by Phytophthora sojae Kaufmann & Gerdemann (Graham et al., 1990; Morris et al., 1991; Graham and Graham, 1994; Graham and Graham, 1996) and to induce nodulation genes in Bradyrhizobium japonicum bacteria that form rhizobia on soybean roots (Kosslak et al., 1987). Isoflavones are also thought to contribute to the positive health effects associated with soybean consumption by humans and animals. Several studies have shown that isoflavones reduce the risk of breast and prostate cancer, cardiovascular disease, and osteoporosis (Messina, 1999; Munro et al., 2003). However, it has been suggested that adverse effects may occur in infants fed soy-based formulae (Setchell et al., 1997). Therefore, breeding soybean cultivars with increased or decreased isoflavone content in the seed may be desirable.

Breeding soybean for desirable individual or total isoflavone content in the seed is difficult since isoflavone content is greatly influenced by the environment. Eldridge and Kwolek (1983) reported that total isoflavone content in soybean seed varied from 1160 to 3090 µg g–1 among four soybean cultivars grown in the same environment and from 460 to 1950 µg g–1 among four Illinois locations for the same cultivar. Wang and Murphy (1994) planted ‘Vinton 81’ over 3 yr at three locations and reported year to have a greater influence on isoflavone content than location. Carrao-Panizzi and Kitamura (1995) tested 22 soybean cultivars at one location in Brazil over 2 yr and attributed the significant differences in total isoflavone content to differences between years in temperature, precipitation and harvest date. Total isoflavone content ranged from 1900 to 3700 µg g–1 for the 100 RILs of the ‘Essex’ x ‘Forrest’ population (Njiti et al., 1999; Meksem et al., 2001; Kassem et al., 2004); however, the authors did not note year effect. Hoeck et al. (2000) tested six soybean cultivars at eight Iowa locations over 2 yr and Lee et al. (2002) planted 15 cultivars at three sites in Korea over 3 yr. In these two studies, genotype, genotype x year, genotype x location, and genotype x year x location interactions were significant for both total and individual isoflavone concentrations.

Several factors have been shown to contribute to environmental effects. Kitamura et al. (1991) observed that high temperatures during the pod filling period significantly decreased isoflavone concentrations in soybean seeds. Tsukamoto et al. (1995) also observed similar effects on soybean plants grown in temperature controlled growth chambers. Vyn et al. (2002) showed that both individual and total isoflavones were positively correlated with leaf and seed K concentrations on low-K soils.

The relationship between isoflavone content and agronomic characteristics is important to breeders. Currently, there is little information on the relationships among these traits. Wang et al. (2000) found that daidzein and genistein content were negatively correlated with yield, days to maturity, and plant height, while total isoflavone content was positively correlated with yield (r = 0.20) among 210 soybean cultivars with a wide genetic range. In contrast, the current study used RIL populations segregating for isoflavone content in an attempt to control background effects. Recently, a mapping study has shown that a QTL for glycitein was closely associated with three seed storage protein genes (Kassem et al., 2004). Furthermore, the mapping population (Essex x Forrest) did not segregate for maturity (less than 7 d) and yet agronomic (yield, height, and lodging) and isoflavone QTLs were correlated (Meksem et al., 2001; Kassem et al., 2004).

The main objective of this study was to determine if isoflavone content was associated with agronomic and seed quality traits in three different RIL populations.


    MATERIALS AND METHODS
 TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Plant Material
On the basis of 3 yr of data (1995–1997), ‘AC756’ (about 1420 µg g–1) and ‘Harovinton’ (about 1560 µg g–1) were used as low isoflavone parents, and ‘OAC Arthur’ (about 1940 µg g–1) and ‘RCAT Angora’ (about 2250 µg g–1) were used as high isoflavone parents (C.-J. Jackson and V. Poysa, personal communication, 2000). The individual and total isoflavone contents in the four parental lines grown in 2002 are presented in Table 1. Three populations were developed for this study. Population 1 (Pop. 1; n = 98) was from the cross Harovinton x OAC Arthur, Pop. 2 (n = 101) from AC756 x OAC Arthur, and Pop. 3 (n = 187) from AC756 x RCAT Angora. In 1999, the F1 plants were self-pollinated in the growth room. The F2 seeds were advanced to F4 by single-seed-descent in the growth room. In 2000, the F4 populations were grown in the field in Harrow, ON, and F4:5 recombinant inbred lines (RILs) established by harvesting individual plants. The RIL populations were sent to a winter nursery in Los Andes, Chile, for a seed increase in November 2000. In 2001, F4:6 seeds were received from Chile and used for planting during the 2001 growing season.


View this table:
[in this window]
[in a new window]
 
Table 1. Isoflavone composition of parental lines used to develop three different RIL populations segregating for isoflavone content. Values are based on a plot mean basis grown at three environments (Harrow, Ridgetown, and Woodstock) in southern Ontario, Canada, in 2002.

 
Seeds of the F4:5 RILs were analyzed for isoflavone content in 2000 and RILs with high, intermediate, and low concentrations of total isoflavone were selected from the three populations for planting in 2002 at three locations (Woodstock, Ridgetown, and Harrow) in southern Ontario, Canada (Fig. 1) . For each RIL population, the following number of RILs were selected from each isoflavone phenotypic class: 10 high, 10 intermediate, and 10 low RILs (n = 30) were selected from Pop. 1; 12 high, 14 intermediate, and 11 low RILs (n = 37) were selected from Pop. 2; 13 high, 11 intermediate, and 14 low RILs (n = 38) were selected from Pop. 3. For Pop. 1 and Pop. 2, RILs that contained <1000, between 1300 and 1400, or >1700 µg g–1 total isoflavones in the seed were classified as being low, intermediate, or, high, respectively. Population 3 was larger and had a mean total isoflavone content different from the other two populations. Therefore, RILs that contained <1000, between 1500 and 1600, or >2100 µg g–1 total isoflavones in the seed were classified as being low, intermediate, or, high, respectively. The four parental lines and 12 maturity checks were also included in the test. Environment Canada weather data for Harrow, Ridgetown, and Woodstock in 2002 is presented in Table 2.



View larger version (49K):
[in this window]
[in a new window]
 
Fig. 1. Frequency distribution of total isoflavone content in soybean seeds for three recombinant inbred line (RIL) populations grown in 2000 (white bars): a) Harovinton x OAC Arthur (n = 98), b) AC756 x OAC Arthur (n = 101), and c) AC756 x RCAT Angora (n = 187). The RILs selected for the 2002 growing season are represented by bars filled with diagonal or horizontal lines. Phenotypic classes selected were RILs containing high, intermediate, and low total isoflavone content in the seed.

 

View this table:
[in this window]
[in a new window]
 
Table 2. Temperature and precipitation normals and monthly average minimum, maximum, and means for the 2002 growing season (Environment Canada, 2002, 2004).

 
An 11 x 11 lattice design with three replications was used at each of the three locations. The plots at Harrow consisted of four rows, 4 m long, with 60 cm between rows. Plots were sprayed with Dual II Magnum (Syngenta Crop Protection Canada, Guelph, ON) at 1.75 L ha–1, Pursuit (BASF Canada, Toronto, ON) at 0.310 L ha–1, and Sencor (Bayer CropScience, Guelph, ON) at 0.6 L ha–1 to control weeds. The soil type at Harrow was Harrow silt loam. The plots at Woodstock consisted of four rows, 5.3 cm long, with 35 cm between rows within the plot and 40 cm between adjacent plots. No border rows were used. Plots were sprayed with Pursuit at 0.312 L ha–1 before planting and on July 2, plots were sprayed with Assure II (Dupont Canada Inc., Mississauga, ON) at 0.63 L ha–1 and Basagran Forte (BASF Canada, Toronto, ON) at 1.75 L ha–1 at 1 L ha–1 to control weeds. The soil type at Woodstock was Guelph loam. In 2002, the field in Woodstock was fertilized with 5-20-20 (N-P-K) applied at 200 kg ha–1. The plots at Ridgetown consisted of five rows 4 m long, with 42 cm between rows. Plots were sprayed with Pursuit at 0.222 L ha–1, Basagran (BASF Canada, Toronto, ON) at 1 L ha–1, and Excel Super (Bayer CropScience, Guelph, ON) at 0.235 L ha–1 to control weeds. The soil type at Ridgetown was clay loam.

Agronomic Data Collection
The following agronomic traits were measured on all plots at each location: seed yield, days to maturity, lodging, and plant height. Seed quality, seed weight, oil content and protein content were measured on an entry mean basis at each location by bulking approximately 150 g of seed from each plot. Seed yield was converted to kg ha–1 adjusted to 130 g kg–1 moisture. Maturity (R8), was recorded when 95% of the pods matured (Fehr et al., 1977). Lodging was scored from 1 (all plants in a plot erect) to 5 (all plants in a plot prostrate). Plant height was estimated as the distance from the soil surface to the tip of the main stem of a representative plant at maturity. Seed size was recorded as the weight of 100 random seeds from a bulk at each location. Seed quality was rated from 1 (seed surface smooth with no evidence of shriveling; disease-free) to 5 (seed very shriveled, cracked seeds, discoloration, evidence of disease). Approximately 300 g of seed, at seed moisture, was used to measure oil and protein contents using near-infrared reflectance (NIR) on a GrainSpec B1126 from FOSS North America Incorporated (Brampton, ON).

Isoflavone Extraction and HPLC Analysis
The soybean seed samples from 2000 were analyzed for isoflavone content on a single plant basis. The samples (about 4 g) were ground in a coffee grinder and sent to University of Guelph Laboratory Services (Guelph, ON) for daidzein, genistein, glycitein, and total isoflavone analysis. Isoflavone concentrations were determined by HPLC as described by Vyn et al. (2002).

Isoflavone analyses of seed samples harvested in 2002 were performed by an alternate, less expensive method developed by Akhtar and Bryan (2002). Briefly, 10 g of soybean seeds from each plot were finely ground in a coffee grinder, with 5 g used for dry matter determination and two grams used for isoflavone extraction. The 2-g samples were mixed with 20 mL of HPLC grade acetonitrile and 4 mL of 0.1 M HCl in a 40-mL screw cap vial. The solutions were heated for 2 h at 60°C in a water bath, and the vials were shaken every 30 min. After cooling, 1 mL aliquots were centrifuged at 21150 x g for 5 min and 100 µL of supernatant and 2 mL of 3N HCl were mixed in a 7-mL glass screw cap vial. The samples were vortexed and hydrolyzed for 24 h at 60°C on a block heater, were allowed to cool, and 2 mL of ethyl ether was added, and the samples were again vortexed. The layers were allowed to separate; the upper ethyl ether layer was removed with a Pasteur pipette, passed through approximately 1 g of sodium sulfate (previously washed with 3 volumes of ethyl ether), and collected in a glass vial. Samples were partitioned with ethyl ether a total of five times, twice with 2 mL and three times with 1 mL. The samples were dried under a gentle stream of nitrogen gas to evaporate the combined ether layers. The dried samples were dissolved in 1 mL of 25% (v/v) acetonitrile, vortexed, and filtered through a 0.45-µm Nylon Cameo 3N syringe filter before HPLC analysis. The Na2SO4 columns were made from 5 3/4 Pasteur pipettes with glass wool used as a plug and approximately 1 g of Na2SO4 used as the column material.

Concentrations of daidzein, genistein, and glycitein were determined by HPLC. The following HPLC apparatus and conditions were used: HPLC, Agilent 1100 binary delivery system from Agilent Technologies Limited (Mississauga, ON) with an auto sampler and a photodiode array (PDA) detector set to collect spectra from 200 to 300 nm; HPLC column, Phenomenex Primesphere C18-HC column (250 x 4.6 mm; 5, 10,15-µm particle size) equipped with a guard cartridge holder (4.0 x 3.0 mm); flow rate, 0.8 mL min–1; and injection volume, 20 µL. HPLC mobile phases were solvent A (10% v/v ACN) and solvent B (38% v/v ACN), and the solvent system was as follows (% solvent A/% solvent B): 0 min (0/100), 5 min (10/90), 20 min (0/100), and 25 min (0/100).

Daidzein, glycitein, and genistein peaks on sample chromatographs were confirmed with standards (Sigma Chemical Company, St. Louis, MO). The Hewlett-Packard software associated with the Agilent HPLC instrument was used to calculate peak areas for each isoflavone. Linear calibration curves were generated for each isoflavone by plotting five known concentrations as a function of peak area. High linearity (R > 0.99) was obtained for each curve. The concentration of each isoflavone (µg per g of sample) was calculated by means of the calibration curves, peak area, sample weight, and dilution factors.

Statistical Analysis
The SAS procedure PROC MIXED (SAS 8.02, SAS Institute Inc., 2001) was used to conduct an analysis of variance for individual and total isoflavone content. Phenotypic class, population, environment, and their interaction were considered fixed effects, and rep(environment) and block(rep) were considered random effects. The class, population and environment means of individual and total isoflavones was calculated by the SAS statement LSMEANS. Comparison of the means for daidzein, glycitein, genistein, and all agronomic and seed quality traits between high, intermediate and low isoflavone phenotypic classes were made by contrasts.


    RESULTS
 TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Distribution
The frequency distributions of the three RIL populations (Fig. 1) for total isoflavone content, on the basis of seed analysis in 2000, departed significantly from normality (P < 0.01). Total isoflavone content for each population showed a peaked (kurtosis = g2) distribution (g2 = +10.3, g2 = +4.1, g2 = +8.0, respectively) that skewed (skewness = g1) toward Harovinton (g1 = +2.2, g1 = +1.4, g1 = +1.7, respectively). The range of values for the RILs exceeded those of the parents for each population. However, the lack of replication in 2000 did not allow for appropriate statistical tests to determine if extreme values were significantly different from parental values.

Population, Environment, and Phenotypic Class Analysis of Isoflavones
There were significant differences among population, environment, and their interaction for individual isoflavone content (Table 3). Population effects were not significant for total isoflavone content, but environment and population x environment interaction were significant. Daidzein, glycitein, and genistein content of Pop. 3 were significantly different from the other populations (Table 4). In general, the mean daidzein and glycitein content were significantly higher and the mean genistein content was significantly lower across RILs for Pop. 3. The mean total isoflavone content of Pop. 3 was higher than the other populations but not significantly different. The mean and range of individual and total isoflavone content for the three populations grown at three locations (Harrow, Ridgetown, and Woodstock, Ontario) in 2002 is presented in Table 5. In general, Ridgetown had the highest and Harrow had the lowest content of daidzein, genistein and total isoflavone for all three populations. A different trend was observed for glycitein. The highest and lowest glycitein contents were found to be in Harrow and Woodstock, respectively, for Pop. 2 and Pop. 3. For Pop. 1, Ridgetown and Harrow had the highest and lowest glycitein content, respectively.


View this table:
[in this window]
[in a new window]
 
Table 3. Significance of the mean squares for population (Harovinton x OAC Arthur, AC756 x OAC Arthur, and AC756 x RCAT Angora), environment (Harrow, Ridgetown, and Woodstock), class ("high", "intermediate", and "low") and their interactions for individual and total isoflavone content of three RIL populations grown at three locations in 2002.

 

View this table:
[in this window]
[in a new window]
 
Table 4. Comparison of overall means, standard errors, and range of daidzein, glycitein, genistein, and total isoflavone content for three RIL populations grown at three environments (Harrow, Ridgetown, and Woodstock, Ontario, Canada) in 2002.

 

View this table:
[in this window]
[in a new window]
 
Table 5. Mean of individual and total isoflavone content of RILs selected from three populations and grown at three environments (Harrow (H), Ridgetown (R), and Woodstock (W), Ontario, Canada) in 2002.

 
Phenotypic classes (high, intermediate, and low) and the interaction (Population x Class) were significantly different for total isoflavone content (Table 3). Mean values for the intermediate isoflavone class were generally close to the midpoint value of the high and low class (Table 6). The interaction effects (Environment x Class) and (Environment x Population x Class) were not significant for total isoflavone content (Table 3).


View this table:
[in this window]
[in a new window]
 
Table 6. Mean and standard error of RILs with "high", "intermediate" and "low" total isoflavone content in three populations across three environments in 2002.

 
Contrasts among the three isoflavone phenotypic classes for daidzein, glycitein, and genistein are shown in Table 7. Differences among the three phenotypic classes were significant for daidzein in Pop. 1 and Pop. 3. As expected, high phenotypic classes had the highest and the low had the lowest daidzein contents. For Pop. 2, the high and intermediate phenotypic classes had daidzein contents significantly higher than the low class, however, the difference between the high and intermediate classes was not significant.


View this table:
[in this window]
[in a new window]
 
Table 7. Contrasts among three classes ("high", "intermediate", and "low") of total isoflavone content for three populations (Harovinton x OAC Arthur, AC756 x OAC Arthur, and AC756 x RCAT Angora) grown at three environments (Harrow, Ridgetown, and Woodstock, Ontario, Canada) in 2002. Data are mean of individual isoflavone content (±SE) (daidzein, glycitein, and genistein) and agronomic traits (±SE) across three levels ("high", "intermediate", and "low") of total isoflavone content.

 
Results for glycitein were not consistent among the three populations (Table 7). The high phenotypic class from Pop. 1 had significantly higher glycitein content than the intermediate and low phenotypic classes. In addition, the intermediate and low phenotypic classes were not significantly different from each other. In Pop. 2, the high and intermediate phenotypic classes had glycitein contents significantly higher than the low class; however, the difference between the high and intermediate classes was not significant. Differences among the three phenotypic classes were not significantly different for glycitein content in Pop. 3.

Results for genistein were similar to daidzein. Differences among the three phenotypic classes were significant for genistein in Pop. 1 and Pop. 2 (Table 7). As expected, high phenotypic classes had the highest and the low had the lowest genistein contents for these two populations. For Pop. 3, the high and intermediate phenotypic classes had genistein contents significantly higher than the low class; however, the difference between the high and intermediate classes was not significant.

There were significant differences among the three isoflavone phenotypic classes for the agronomic traits yield, maturity, plant height, and lodging. The significant differences were not consistent among the three populations (Table 7). Yield, maturity, and plant height differed significantly in Pop. 1 and Pop. 2. The high and intermediate phenotypic classes produced significantly more seed, matured significantly later, and were significantly taller than low phenotypic classes (Table 7). Furthermore, differences between high and intermediate classes in the two populations were not significant for these three traits. As in the two other populations, the high phenotypic class matured significantly later than the low phenotypic class for Pop. 3. However, the difference in maturity between the low and intermediate phenotypic classes was not significant for this population. In Pop. 1, lodging was significantly different among the three phenotypic classes. The high phenotypic class lodged the most and the low lodged the least. In Pop. 2, none of the tests were significant. Interestingly, the high and low classes were significantly different from the intermediate class but were not significantly different from each other in Pop. 3.

Contrasts for protein content, oil content, and seed quality for the three populations are given in Table 7. Differences in protein content, oil content, and seed quality for Pop. 2 and Pop. 3 were not significant for a large number of the contrasts. However, there were highly significant differences for Pop. 1. The high and intermediate phenotypic classes had significantly lower protein content and significantly better seed quality than low phenotypic classes (Table 7). Furthermore, differences between high and intermediate classes were not significant for these two traits. Seed weight differences among isoflavone phenotypic classes were not significant in any of the populations.


    DISCUSSION
 TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Daidzein, glycitein, and genistein contents were significantly different among the three populations (Table 4). It was not surprising to find that Pop. 1 and Pop. 2 favored the synthesis of genistein since these populations were derived from OAC Arthur, a cultivar with high genistein content (Table 1). Similarly, Pop. 3 favored the synthesis of daidzein and glycitein (Table 4) probably because of the higher daidzein and glycitein content in RCAT Angora in comparison to OAC Arthur (Table 1). Differences in isoflavone content among the three populations suggested that different genetic backgrounds might cause isoflavone alleles to be expressed differently and/or that different alleles were segregating in the populations. Other studies have also reported significant differences in isoflavone content because of genotype (Njiti et al., 1999; Hoeck et al., 2000; Lee et al., 2002). Therefore, by combining favorable alleles one could develop soybean seeds with desirable isoflavone content.

Breeding soybean for high or low isoflavone content in the seed is challenging since isoflavone content is influenced by environment (Eldridge and Kwolek, 1983; Wang and Murphy, 1994; Carrao-Panizzi and Kitamura, 1995; Hoeck et al., 2000; Lee et al., 2002). In this study, the effects of environment on isoflavone content were also evident as there were significant differences among the three locations, Harrow, Ridgetown, and Woodstock, in 2002 (Table 3 and 5). Differences in isoflavone content among the three locations could have been partly attributed to differences in temperature and precipitation since soybean RILs grown in Harrow, a location that had higher temperatures and received the least amount of precipitation during the field season (Table 2) in comparison to Ridgetown and Woodstock, had the lowest concentrations of daidzein, genistein, and total isoflavone. The results supported the conclusion of Kitamura et al. (1991) and Tsukamoto et al. (1995) that high temperatures during the pod filling period significantly decreased isoflavone concentrations in soybean seeds. Environmental factors such as planting date, cultural practices, soil fertility (Vyn et al., 2002), or weed, disease, or insect pressure also could affect isoflavone content but these factors were not investigated in our study.

Although environment had a significant effect on total isoflavone content (Table 3 and 5), phenotypic classification of the RILs did not change significantly. In fact, total isoflavone phenotypic groups remained significantly different for each population (Table 6). It was also noted that very few of the RIL phenotypic classifications changed when the 2000 and 2002 isoflavone contents were compared (data not shown), confirming the validity of the second method of isoflavone analysis. Other studies have also shown relatively consistent ranking of soybean genotypes with different isoflavone content when tested in multiple environments. Lee et al. (2002) found that the genotype ‘Geomjeong’ had consistently higher total isoflavone content across 3 yr and three locations. Eldridge and Kwolek (1983) found that ‘Corsoy 79’ had greater isoflavone contents than ‘Hardin’ at each of four locations. Chiari et al. (2004) and Meksem et al. (2001) reported high heritability estimates for isoflavone content. Therefore, results from this study and previous research are encouraging to breeders because it should be possible to select soybean lines with relatively high or low isoflavone content in the seeds when tested in multiple environments.

The three main isoflavones found in soybean seed, genistein, daidzein, and glycitein, are generally in the concentration ratio 1:1:0.2 (Manach et al., 2004). Since daidzein and genistein account for a majority of the total isoflavone content, it was not unusual to find that contrasts among total isoflavone phenotypic classes for daidzein and genistein were significant in the three populations and that their relationships were linear. Glycitein, on the other hand, does not contribute very much to the total isoflavone content, which resulted in fewer significant contrasts. In fact, there were no significant contrasts among the three phenotypic classes for Pop. 3. This supported the findings of Primomo et al. (2005), who showed that QTL for total isoflavone mapped to similar regions as daidzein and genistein but not glycitein. The results suggested that one could simultaneously breed soybean lines for daidzein and genistein but not for glycitein when breeding for total isoflavone content.

Information on agronomic performance of soybean cultivars with diverse isoflavone content is limited. Wang et al. (2000) reported no significant association between total isoflavone content in soybean seeds and days to maturity or plant height, but they did find a weak, positive correlation with seed yield. There may be two possible reasons for the different results obtained between the two studies. First, RIL populations, lines similar by descent, were used in our study, whereas Wang et al. (2000) used 210 assumedly unrelated soybean cultivars. Second, our study analyzed soybean seeds for the three main aglycones (i.e., daidzein, genistein, and glycitein), whereas Wang et al. (2000) analyzed them for the 12 isoflavone components, the aglycones, and their corresponding glucosides.

In general, soybean cultivars late in maturity have higher yields, are taller, and lodge more than early maturing cultivars. In this study, high and low phenotypic classes were significantly different for maturity in all three populations (Table 7). Hence, it was not surprising to find that the high phenotypic class produced more seeds, were taller, and lodged more than the low in Pop. 1 and Pop. 2. Interestingly, the differences in maturity between the two groups were 10 and 3 d for Pop. 1 and Pop. 2, respectively. Clearly, isoflavone content had a positive effect on agronomic traits. Even though Pop. 2 had very limited segregation in maturity, it was possible to develop individual RILs with high seed yields and desirable isoflavone content. RILs with the highest and lowest total isoflavone content contained 2030 and 970 µg g–1 and had seed yields of 4475 and 4222 kg ha–1, respectively, when averaged across the three environments (data not shown). Our results support a mapping study that showed a QTL for daidzein content was closely linked to a seed yield QTL in the mapping population Essex x Forrest (Meksem et al., 2001).

Significant differences among the three phenotypic classes were detected for seed quality, oil content, and protein content (Table 7). The lack of a consistent trend, in addition to the minimal differences between the phenotypic classes, suggested that isoflavone content in the seed likely does not have an effect on seed quality and oil content. It was apparent that isoflavone content did not have an effect on seed weight because phenotypic classes were not significantly different for any of the three populations.

In a recent genetic study, Chiari et al. (2004) reported a negative correlation between isoflavone and protein content, ranging from –0.51 to –0.37. In our study, the negative relationship reported by Chiari et al. (2004) was also observed in Pop. 1; however, it was not evident in Pop. 2 and Pop. 3. Two RILs with high isoflavone content, 1851 and 1746 µg g–1, also contained high protein content, 431 and 442 g kg–1, respectively. This is very encouraging to breeders, because it suggests that high soybean protein is compatible with high isoflavone content, which could be advantageous for the functional soy food industry.

The negative correlation between isoflavone and protein content may be attributed to the result of gene linkage, pleiotropy, or environmental conditions during seed maturation. Recently, two independent mapping studies using different genetic material (Kassem et al., 2004; Primomo et al., 2005) showed that QTL for glycitein and protein content map to common regions of the soybean genome, supporting the hypothesis of genetic linkage.

In conclusion, this study has provided an evaluation of the agronomic performance and seed quality of RILs segregating for isoflavone content in soybean seeds. Our results have indicated that it was possible to develop RILs with high yields and low isoflavone content, despite earlier reports that later maturing cultivars had higher isoflavone content. Similarly, it was possible to develop RILs with high isoflavone and high protein content despite earlier reports of a negative correlation between these two traits. Using traditional breeding methods to select soybean with desirable isoflavone content in the seed requires testing in multiple environments because of the substantial environmental variation for this trait. In addition, applying early selection when breeding for isoflavone content in soybean seeds might not be practical since isoflavone analysis of each RIL could be very expensive. Thus, the use of marker-assisted selection is advantageous in the design of an efficient and cost-effective breeding strategy for developing soybean cultivars with desirable isoflavone content in the seed. Despite a very limited range of maturity in Pop. 2 and Pop. 3, high, intermediate, and low isoflavone RILs were identified with no significant difference for yield, height, and lodging. Clearly, these two populations would be superior for mapping isoflavone and agronomic traits because the maturity effect is removed.


    ACKNOWLEDGMENTS
 
The technical assistance of Wade Montminy, Bryan Sterling, Bob Armstrong, Cal Klager, Yesenia Salazar, Claudia Schmidt, Vicky Godfrey, Rosanna Vallejos, Claudio Sallese, and Julia Zilka is greatly appreciated. Drs. Irena Rajcan and Brian Allen are acknowledged for their helpful suggestions regarding the statistical analysis. We thank Dr. M.H. Akhtar and Mike Bryan of Agriculture and Agri-Food Canada, Guelph, ON, for kindly providing their time and lab facilities for isoflavone extraction and HPLC analysis. Appreciation is extended to the Ontario Ministry of Agriculture and Food (OMAF), Ontario Soybean Growers, Canadian Fertilizer Institute, First Line Seeds, the Potash and Phosphate Institute of Canada, Hyland Seeds, Agricultural Adaptation Council of Canada, and Hensall District Co-op for the financial support to this research.


    NOTES
 TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
This work was submitted by Valerio S. Primomo in partial fulfillment for the Ph.D. degree from the Dep. of Plant Agriculture, Crop Science Division at the Univ. of Guelph.

Received for publication October 19, 2004.


    REFERENCES
 TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
Crop Sci.Home page
S. E. Murphy, E. A. Lee, L. Woodrow, P. Seguin, J. Kumar, I. Rajcan, and G. R. Ablett
Genotype x Environment Interaction and Stability for Isoflavone Content in Soybean
Crop Sci., June 26, 2009; 49(4): 1313 - 1321.
[Abstract] [Full Text] [PDF]


Home page
Crop Sci.Home page
M. J. Morrison, E. R. Cober, M. F. Saleem, N. B. McLaughlin, J. Fregeau-Reid, B. L. Ma, W. Yan, and L. Woodrow
Changes in Isoflavone Concentration with 58 Years of Genetic Improvement of Short-Season Soybean Cultivars in Canada
Crop Sci., November 24, 2008; 48(6): 2201 - 2208.
[Abstract] [Full Text] [PDF]


This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Primomo, V. S.
Right arrow Articles by Rajcan, I.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Primomo, V. S.
Right arrow Articles by Rajcan, I.
Agricola
Right arrow Articles by Primomo, V. S.
Right arrow Articles by Rajcan, I.
Related Collections
Right arrow Soybean
Right arrow Plant and Environment Interactions
Right arrow Crop Genetics


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education
Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality
The Plant Genome