Crop Science 43:511-514 (2003)
© 2003 Crop Science Society of America
CROP BREEDING, GENETICS & CYTOLOGY
Genotype and Environment Influence on Protein Components of Soybean
Walter R. Fehr*,a,
Joseph A. Hoecka,
Susan L. Johnsona,
Patricia A. Murphyb,
Joel D. Nottc,
Gabe I. Padillaa and
Grace A. Welkea
a Dep. of Agronomy, Iowa State Univ., Ames, IA 50011
b Dep. of Food Science and Human Nutrition, Iowa State Univ., Ames, IA 50011
c Protein Facility, Iowa State Univ., Ames, IA 50011
* Corresponding author (wfehr{at}iastate.edu)
 |
ABSTRACT
|
|---|
The quality of some soybean [Glycine max (L.) Merr.] food products has been associated with the content of some components of the protein fraction of the seed. The objective of this study was to determine the role of genotype, environment, and genotype x environment interactions on the components of ß-conglycinin (ßc) and glycinin (G). The traits were measured on 14 cultivars of Maturity Group II grown in three replications at each of eight locations throughout Iowa in 1998, 1999, and 2000. Cultivars were significantly different for all traits, except the A3 subunit of G. No significant interactions were found for cultivars with years or locations for ßc, G, or the G/ßc ratio, which indicated that the relative performance of cultivars across environments was consistent. No significant differences were expressed among years or locations for ßc, G, and the G/ßc ratio; however, there were significant differences for the three traits among the 24 environments. There was a significant phenotypic correlation of -0.92 between the ßc and G contents of the 14 cultivars averaged across environments, but no significant correlations of ßc, G, or the G/ßc ratio with protein and oil content. It should be possible to breed for desired levels of the protein components in a cultivar development program and to select among cultivars for the traits in commercial grain production.
|
INTRODUCTION
|
|---|
ß-CONGLYCININ and glycinin are the two primary components of soybean storage protein (Wilson, 1987). The components of ßc that can be separated by electrophoresis are 
, ', and ß peptides. The components of G can be separated into A3; total A1a, A1b, A2, A4; and total basic peptides. Differences in ßc, G, and the G/ßc ratio have been shown to influence the quality of soyfood products (Murphy et al., 1997; Cai and Chang, 1999; Tezuka et al., 2000).
Significant differences among genotypes and environments have been reported for ßc, G, and the G/ßc ratio. Murphy and Resurreccion (1984) evaluated 10 soybean cultivars grown at Ames, IA, during 1980 and 1981 and observed significant variation among cultivars for ßc, G, and the G/ßc ratio. They found a significant difference between years for G, but not for ßc. Helms et al. (1998) grew two soybean cultivars at four locations in Minnesota and North Dakota during 1993 and five locations in the two states during 1994. The G/ßc ratio for the two cultivars was not significantly different in eight of the nine environments. They observed significant differences among locations within years for the G/ßc ratio.
The studies of Murphy and Ressurreccion (1984) and Helms et al. (1998) did not provide a complete analysis of the impact of genotype, year, location, and their interactions on the protein components because multiple cultivars were not grown at the same locations during several years. The objective of our study was to examine the influence of genotype, environment, and genotype x environment interactions on the components of ßc and G.
|
MATERIALS AND METHODS
|
|---|
The 14 cultivars of Maturity Group II used in this experiment were developed by Iowa State University. They represent the range of general-use and special-use cultivars grown in the midwestern United States. IA2021 is a general-use cultivar; IA2040 is a large-seeded cultivar; HP204, IA2017, IA2020, IA2034, IA2041, IA2042, and Vinton 81 are large-seeded, high-protein cultivars; IA2035 is a small-seeded cultivar; and IA2025, IA2027, IA2029, and IA2032 are lipoxygenase-free cultivars. The cultivars were grown in three replications of a randomized complete-block design at Greene, Sioux Rapids, Kanawha, Hubbard, Grand Junction, Winterset, Griswold, and Crawfordsville, IA, in 1998, 1999, and 2000. The soil types are a Kenyon silty clay loam (fine-loamy, mixed, mesic Typic Hapludoll) at Greene, a Primghar silty clay loam (fine-silty, mixed, mesic Aquic Hapludoll) at Sioux Rapids, a Webster silty clay loam (fine-loamy, mixed, mesic Typic Haplaquoll) at Kanawha, a Nicollet loam (fine-loamy, mixed, mesic Aquic Hapludoll) at Grand Junction and Hubbard, a Winterset silty clay loam (fine, montmorillonitic, mesic Typic Argiudoll) at Winterset, a Marshall silty clay loam (fine-silty, mixed, mesic Typic Hapludoll) at Griswold, and a Mahaska silty clay loam (fine, montmorillonitic, mesic Aquic Argiudoll) at Crawfordsville. The plots consisted of four rows 3.1 m long spaced 0.69 m apart and planted at a seeding rate of 30 seeds m-1. The entire length of the center two rows was harvested with a self-propelled combine.
The 72 replications in the experiment were randomized, and the plots of cultivars in a replication were analyzed for the protein components in order from the lowest to the highest plot number. For each plot, enough whole soybean seeds were ground with an electric household coffee grinder to obtain a 25-g sample. The 25-g ground sample was reconstituted with 250 mL deionized water. The sample was stirred as the pH was adjusted to 8.6 with 6 M NaOH. The samples were stored at 5°C overnight (Wu et al., 2000). A 25-mL aliquot of the resulting supernatant was decanted into a 50-mL centrifuge tube, and the sample was centrifuged at 15000 x g for 20 min at 20°C. A 50-µL aliquot of the supernatant was added to 50 µL of extraction buffer [50 mM tris (hydroxymethyl) aminomethane (THAM), pH 8.0; 5.0 M urea; 0.2% sodium dodecyl sulfate (SDS) (w/v); 2% 2-mercaptoethanol (v/v)], and incubated for 2 h at room temperature. A 100-µL aliquot of 2X sample buffer [125 mM THAM, pH 6.8; 5.0 M urea; 0.2% SDS (w/v); 2% 2-mercaptoethanol (v/v); 20% glycerol (v/v); 0.01% bromophenol blue] was added to the tube and heated at 100°C for 10 min.
Sodium dodecyl polyacrylamide gel electrophoresis (SDS-PAGE) was performed by the procedure of Laemmli (1970), with 12.5% (w/v) separating gels and 4% (w/v) stacking gels. Of the 15 wells in the gel, 14 were used for test samples and 1 outside well was used for a control sample. The control sample used for all gels was prepared from seed harvested from one plot of IA2021 grown at Ames, IA, in 2000. A control sample was necessary to account for differences in band intensities among gels caused by variation in staining and destaining. A 2-µL sample of extracted protein was loaded into each well. The gels were run in a buffer solution described by Yaklich (2001) at 180V for 45 min and stained with Coomassie Blue for 40 min. The gels were destained for 2 h, the solution was changed, and destaining continued for another hour.
The protein components measured as bands on the electrophoresis gels were ; '; ß; A3; total A1a, A1b, A2, A4; and basics. Total A1a, A1b, A2, A4 was one band on a gel. Basics were two adjacent bands measured separately (Cai and Chang, 1999). The protein bands on the destained gels were quantified with the Kodak 1D Image Analysis Version 3.5 software package (Eastman Kodak Company, Rochester, NY). The amount of a protein component in a test sample was expressed as a percentage of the control sample in the same gel. The area of each protein band of a test sample was divided by the area of the same protein band for the control sample on the gel, and the quotient was multiplied by 100. For ßc, the percentages for ', , and ß were added together, and the sum was divided by three to obtain the mean of the subunits. Total acidics was the sum of the percentages for A3 and total A1a, A1b, A2, A4 divided by two. For total basics, the percentages of the two adjacent bands designated basics were added together and the sum was divided by two. G was the sum of the percentages of total acidics and total basics divided by two. The G/ßc ratio was computed as the percentage of G divided by the percentage of ßc.
For the analysis of all traits, genotypes were considered fixed effects, and replications, years, and locations were considered random effects. One combined analysis of variance was computed to estimate the main effects of the 3 yr, eight locations, 14 cultivars, and their interactions. A second combined analysis of variance was computed to estimate the main effects of the 14 cultivars and 24 environments and the cultivar x environment interactions. The analyses of variance were performed with the general linear model (GLM) procedure of the SAS software (release 8.0) (SAS Institute, 2000). The significance of the mean squares for the main effects and interactions was calculated by deriving the expected mean squares and determining the appropriate F tests. Phenotypic correlation coefficients were computed among traits of the 14 cultivars on the basis of their means averaged across years and locations. The correlations were calculated by the correlation (CORR) procedure of the SAS software (release 8.0) (SAS Institute, 2000).
|
RESULTS AND DISCUSSION
|
|---|
There were significant differences among cultivars for all protein components, except A3 (Table 1). The range among cultivars for A3 was only 8 percentage units compared with a range of at least 16 percentage units for the other components. There was a strong inverse relationship between the ßc and G content of the cultivars with a significant (P < 0.01) phenotypic correlation between the traits of -0.92. The only other study that has examined the relationship between the ßc and G content of multiple cultivars was that of Murphy and Resurreccion (1984). The data they provided were used to compute the phenotypic correlation between ßc and G on the basis of the average content of their 10 cultivars grown in 2 yr at Ames, IA. Their phenotypic correlation between the two traits of -0.22 was not significant (P > 0.05). The difference in the correlation between ßc and G in the two studies may be due to the cultivars studied and the number of environments used to evaluate their performance. Cultivar performance for ßc and G was based on 24 environments in our study compared with only 2 environments for Murphy and Resurreccion (1984).
View this table:
[in this window]
[in a new window]
|
Table 1. Mean content and rank of 14 soybean cultivars for protein components averaged across 3 yr and eight Iowa locations.
|
|
The cultivars differed significantly for the G/ßc ratio. The range was from 2.04 for Vinton 81 to 1.51 for IA2021 (Table 1). The high ranking for Vinton 81 was similar to that observed by Murphy and Resurreccion (1984) for ‘Vinton’, the recurrent parent used for developing Vinton 81 by backcrossing. In their study, Vinton had the highest G/ßc ratio of the 10 cultivars they evaluated. Vinton 81 is considered an excellent cultivar for the production of tofu, which may be due in part to its consistently high G/ßc ratio compared with other available cultivars.
The ßc content of the cultivars had a significant negative phenotypic correlation of -0.99 (P < 0.01) with the G/ßc ratio, and the G content had a significant positive correlation of 0.96 with the ratio. The large coefficients reflect the strong negative correlation between ßc and G. With the data provided by Murphy and Resurreccion (1984), the phenotypic correlations of ßc and G with the G/ßc ratio were computed for the 10 cultivars they studied. The G/ßc ratios had a significant negative coefficient of -0.81 with ßc and a significant positive coefficient of 0.74 with G.
The percentages of ßc and G and the G/ßc ratio were not significantly correlated with the protein and oil content of the cultivars. The phenotypic correlation coefficients with protein content were -0.34 for ßc, 0.37 for G, and 0.33 for the G/ßc ratio. The correlation coefficients with oil content were 0.20 for ßc, -0.08 for G, and -0.15 for the G/ßc ratio. The lack of significant correlations indicated that it should be possible to select for desired contents of ßc and G among cultivars with similar protein contents. This would be advantageous for production of soybean foods that all require high-protein cultivars, but for which different ßc and G contents would be desirable.
No significant differences were found among years for the protein components, except for ' and ß (Table 2). No significant differences were observed among locations for any of the components. The year x location interaction was only significant for ß and total acidics. When the analyses of variance were computed as 24 environments without regard to year and location, significant differences (P < 0.05) among environments were observed for all traits. The results indicated that ßc, G, and the G/ßc ratio were influenced by the environment, but that major changes in the three traits should not be expected from one year to the next if harvested soybean seed from similar locations is mixed together for manufacturing a food product. The lack of significant differences among the eight locations indicated that no site within the test area would be expected to be consistently different from another for ßc, G, and the G/ßc ratio. Differences would be expected among year–location combinations (environments), which could be important if seed from individual fields was kept separate for food manufacturing, instead of mixing seed from multiple fields grown in the same or different years.
The significant influence of environment on ßc, G, and the G/ßc ratio observed in our study was in agreement with the results of Murphy and Resurreccion (1984) and Helms et al. (1998). Our results did not support the conclusion of Murphy and Resurreccion (1984) that environmental differences were more important than cultivar differences for ßc and G. They based their conclusion on the analysis of four cultivars, each of which was grown in different years and locations. Cultivar differences could not be assessed in that situation because they were not grown in a common environment. When they analyzed seed of 10 cultivars grown in the same environments at Ames, IA, the mean G content was 50.3% in 1980 and 51.6% in 1981. The range of G among cultivars in 1980 was from 46.9 to 54.4% and in 1981 was from 46.7 to 57.2%. There also was a greater range among cultivars than years for ßc. Cultivar selection would be an important consideration if a particular ßc, G, or G/ßc ratio were important for the manufacture of a food product.
The relative performance of the 14 cultivars was consistent among years and locations. The cultivar x year interaction was only significant for total A1a, A1b, A2, A4. The cultivar x location interaction was not significant for any of the traits. When the data were analyzed as 24 environments without regard to year or location, the cultivar x environment interaction was not significant for any of the protein components or the G/ßc ratio. It should be possible to assess the relative performance of cultivars for ßc, G, and G/ßc with a limited number of test environments. For commercial soybean production, cultivars could be selected for preferred ßc, G, or G/ßc composition.
|
NOTES
|
|---|
Journal Paper No. J-19771 of the Iowa Agric. and Home Econ. Exp. Stn., Ames, IA. Project No. 3732 and supported by the Hatch Act, State of Iowa, Iowa Soybean Promotion Board, and Raymond F. Baker Center for Plant Breeding.
Received for publication March 2, 2002.
|
REFERENCES
|
|---|
- Cai, T., and K. Chang. 1999. Processing effect on soybean storage proteins and their relationship with tofu quality. J. Agric. Food Chem. 47:720–727.[Medline]
- Helms, T.C., T.D. Cai, K.C. Chang, and J.W. Enz. 1998. Tofu characteristics influenced by soybean crop year and location. http://www.ag.ndsu.nodak.edu/ndagres/ndagres.htm; verified 25 October 2002.
- Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680.[Medline]
- Murphy, P.A., H. Chen, C.C. Hauck, and L.A. Wilson. 1997. Soybean protein composition and tofu quality. Food Tech. 51:86–88, 110.
- Murphy, P.A., and A.P. Resurreccion. 1984. Varietal and environmental differences in soybean glycinin and ß-conglycinin content. J. Agric. Food Chem. 32:911–915.
- SAS Institute. 2000. SAS guide for personal computers. 8th ed. SAS Inst., Cary, NC.
- Tezuka, M., H. Tiara, Y. Igarashi, K. Yagasaki, and T. Ono. 2000. Properties of tofus and soy milks prepared from soybean having different subunits of glycinin. J. Agric. Food Chem. 48:1111–1117.[Medline]
- Yaklich, R.W. 2001. ß-conglycinin and glycinin in high-protein soybean seeds. J. Agric. Food Chem. 49:729–735.[ISI][Medline]
- Wilson, R. 1987. Seed metabolism. p. 643–686. In J.R. Wilcox (ed.) Soybeans: Improvement, production, and uses. ASA, CSSA, and SSSA, Madison WI.
- Wu, S., P.A. Murphy, L.A. Johnson, M.A. Reuber, and A.R. Fratzke. 2000. Simplified process for soybean glycinin and ß-conglycinin fractionation. J. Agric. Food Chem. 48:2702–2708.[Medline]
This article has been cited by other articles:
|
|
|
|
 
M. E. Stoll, W. T. Schapaugh Jr., M. S. Zutara, J. Hu, and X. S. Sun
Genotype and Environment Effects on Adhesive Shear Strength in Soybean-Based Adhesives
Crop Sci.,
September 8, 2006;
46(5):
2008 - 2012.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
