Nakasathien Et Al.: Nitrogen Supply and Seed Protein Concentration in Soybean

nificant barriers to development of high-protein commercial cultivars. The physiological and biochemical basis for increased seed protein Soybean SPC is inherited as a quantitative trait (Burconcentrations (SPC) observed in restriction-index, recurrent-selecton, 1987) and influenced by environmental effects tion breeding programs with soybean [Glycine max (L.) Merr.] are poorly understood. The hypothesis that soybean SPC is regulated by (Burton, 1988). Generally the trait is much less influthe supply of nitrogenous substrates available to the seed was evaluenced by the genotype of the embryo than by the genoated. Effects of supra-optimal external N on seed storage protein type on which the seeds develop (Singh and Hadley, accumulation, amino acid concentration and composition in leaves 1968). This suggests that whole plant processes such as and seeds at R5, and levels of specific storage protein subunits were N acquisition, translocation, and mobilization of C and measured. Genotypes with different SPC (NC 107, normal; N87-984N are important in the determination of seed protein 16, intermediate; and NC 111, high) were grown in controlled-environconcentration. ment chambers and supplied with 30 mM N as NH4NO3 from V5 to Brim and Burton (1979) used recurrent selection to maturity or from R5 to maturity. Control plants received 10 mM N increase SPCs in two populations (IA and IIA). Carter throughout the growth cycle. Relative to control, supra-optimal N et al. (1982) examined relationships between N accumuincreased SPC of NC 107 and N87-984-16 by an average of 28%. Greater enhancement of protein accumulation than of dry matter lation and distribution and SPC in high and low seed accumulation in the seed resulted in SPCs of 460 to 470 g kg21, germplasm from both populations. They observed that which are appreciably greater than concentrations observed for these high protein germplasm from population IA accumucultivars grown in the field. Supra-optimal N also increased SPC of lated more total N before reproductive development the high protein line (NC 111) by 15%, but this increase resulted than high seed protein germplasm in population IIA. entirely from a decrease in yield. Supra-optimal N supplied to NC Selected high and low seed protein lines derived from 107 and N87-984-16 from V5 until R5 increased total free amino acid Cycle 0 and advanced cycles of selection for both popuconcentrations in seeds and leaves at R5 by an average of 21 and lations were evaluated for vegetative N accumulation 46%, respectively. Enhanced accumulation of the b subunit of b prior to reproductive growth and vegetative N mobilizaconglycinin which does not contain methionine and cysteine accounted tion to seed during reproductive growth (Burton et al., for the increase in SPC. While enhanced N availability increased the SPC of a normal protein line into the high range, availability of sulfur 1995). The authors concluded that the high protein conamino acids in the developing seed determined which storage protein centration trait resulted neither from greater vegetative subunits were synthesized from the extra N. N accumulation before R5 nor from greater N mobilization from vegetative tissue to developing seed since normal and high seed protein lines did not differ signifiW considerable variation in SPC exists within cantly in these attributes. Thus, measurements of whole soybean germplasm (350–500 g kg21 dry weight), plant N accumulation and vegetative N mobilization it has been difficult to enhance this trait in soybean seed have not provided a clear understanding of how SPC without lowering seed yield and oil concentration (Brim in soybean is regulated. and Burton, 1979; Miller and Fehr, 1979). These negaSaravitz and Raper (1995) evaluated the C and N tive impacts on yield and constituent value remain sigrequirements of ‘Ransom’ soybean embryos in in vitro culture from 17 to 41 d after flowering (DAF). When 150 mM sucrose was used as the C source, the protein S. Nakasathien, R.F. Wilson, and P. Kwanyuen, USDA-ARS, Dep. concentration in the embryo increased from 150 to 690 of Crop Science; and D.W. Israel, USDA-ARS, Dep. of Soil Science, g kg21 dry weight as the glutamine concentration was North Carolina State University, Raleigh, NC 27695-7620. A cooperaincreased from 0.6 to 120 mM. The protein concentrative investigation of the USDA-ARS, the North Carolina Agricultural Research Service and U.S. Soybean Grant, project 8227. From the tion at the highest glutamine concentration (690 g kg21 dissertation of the senior author in partial fulfillment of requirements for a Ph.D. degree at North Carolina State University. Received 3 Abbreviations: DAF, days after flowering; DAT, days after transplant; June 1999. *Corresponding author ([email protected]). SPC, seed protein concentration; YFEL, youngest fully expanded leaves. Published in Crop Sci. 40:1277–1284 (2000). 1278 CROP SCIENCE, VOL. 40, SEPTEMBER–OCTOBER 2000 Table 1. Daily N application for each treatment at different stages of the experiment to initiate reproductive development. Day of growth. and night temperature of 26 and 228C, respectively, were used throughout the growth cycle. Treatments Growth interval Control 30 mM N after V5 30 mM N after R5 Nutritional Treatments DAT† mmol N d21 4–20 (V5)‡ 2.5 2.5 2.5 With the exception of inorganic N, the composition of nutri20–76 (R5)§ 5.0 15.0 5.0 ent solutions was the same as that previously described by 76–maturity¶ 10.0 30.0 30.0 McClure and Israel (1979). From the day of transplanting until 3 DAT, noninoculated plants were supplied only 0.25 L of † Days after transplanting. ‡ Vegetative stage when plants have six nodes with leaves with undeionized water. From 4 DAT until the N treatments were folded leaflets. imposed, pots were flushed with 1.0 L of deionized water at § Beginning seed fill stage when plants have pods at uppermost four nodes 0800 and 1300 h and supplied with nutrient solution as shown on main stem with seeds 3 mm in length. ¶ After R5 stage, 0.5 L of nutrient solution was applied after both morning in Table 1. and afternoon flushes with deionized water, as plants in the control Since the pH of the root zone is altered by the form of N treatment (10 mM N) showed a moderate N deficiency with a single being absorbed, the pH of the bulk solution within the solid daily application. substrate was measured twice weekly by collecting the first 0.5 L of solution that drained from the bottom of the pots dry wt.) is much higher than that (500 or 550 g kg21 dry during flushing with deionized water in the morning. The ratio wt.) considered to be high for soybean seeds produced of NH4 and NO3 was adjusted to maintain the pH of the root on an intact plant. When grown under field conditions, zone within the range of 5 to 7. When the desired NH4NO3 the Ransom cultivar has SPC of 380 to 400 g kg21 dry ratio was ,1.0, KNO3 was substituted for the appropriate amount of NH4NO3. weight. In this system, the embryos developing in high glutamine medium could be considered to have an unlimited supply of N available to support protein biosynSampling Procedures thesis. The dry mass of embryos also doubled as the Four replicates of the control and N-treated plants were protein concentration increased with increased glutaharvested at R5 and maturity. Plants were separated into mine supply (Saravitz and Raper, 1995). Therefore, releaves, stem, roots, seeds, and pod walls and oven-dried at striction of growth by glutamine did not cause the in608C for 72 h. Dry mass and N concentration of plant parts crease in protein concentration. These results imply that and whole plants were measured. soybean seeds of this normal seed protein cultivar have intrinsic biochemical capacity to synthesize high protein Tissue Nitrogen and Sulfur Analysis concentrations and that N available to the developing Samples (50–100 mg) of each plant part were digested by a seed may regulate SPC. Kjeldahl procedure using a zirconium copper catalyst (Glowa, The objective of our research was to test the hypothe1974) and a salicylic acid predigestion step to convert NO3 to sis that high SPC trait is regulated by the supply of NH4 (Nelson and Sommers, 1973). Digests were alkalinized nitrogenous substrates to the developing seeds by meawith 10 M NaOH, and NH3 was steam-distilled into 0.32 M suring the effects of supra-optimal external N supply boric acid and quantified by titration with 0.01 M potassium biiodate. Nitrogen percentage in seed was converted to protein on seed protein formation and associated physiological percentage by multiplying by the conversion factor, 6.25. Suland biochemical processes in genotypes that exhibit norfur concentration in seed was determined by inductively coumal and high SPC. If N supplied by the plant limits SPC, pled plasma emission spectrometry after sample digestion in supply of supra-optimal N to normal protein lines is concentrated nitric acid (Novozamsky et al., 1986). predicted to increase SPC. Amino Acid Analysis MATERIALS AND METHODS At R5, leaves, seeds, and pod wall were subsampled for Plant Culture amino acid analysis. Six leaf-discs (total area of 2.65 cm) were collected from the youngest fully expanded leaves (YFEL). Normal (NC 107) and high (NC 111) protein lines derived from recurrent selection populations (Carter et al., 1986) and Five-hundred milligrams of seeds and pod walls were subsampled and cut into small pieces. These plant parts were frozen N 87-984-16, an intermediate protein line derived from restriction-index recurrent selection, were used for this study (Burin liquid N and placed in a freezer at 2808C until the analysis was performed. ton, 1988). Seeds were pregerminated in germination paper saturated with 0.5 mM CaSO4 at 288C and 95% relative humidApproximately 0.5, 0.5, and 0.05 g fresh weight of seeds, pod wall, and leaf were homogenized in 13.7 M ethanol using ity for 72 h before transplanting into 25.4-cm pots filled with coarse perlite. Two to three seedlings were transplanted in the Brinkman Polytron (Brinkman Instruments, Westbury, NY) for extraction of soluble N. Homogenates were centrieach pot and thinned to one plant per pot at 14 d after transplanting (DAT). Plants were grown in two controlled-environfuged at 1240 g in a swinging bucket centrifuge (Model CR411, Jouan Inc., East Winchester, VA). The pellets were extracted ment chambers with two replications in each chamber in the Southeastern Plant Environment Laboratories at North Carotwo more times by suspending in 13.7 M ethanol followed by centrifugation. Three ethanolic supernatant fractions conlina State University. The photosynthetic photon flux density of ≈850 to 900 mmol m2 s2 between wavelengths of 400 to taining soluble N were combined. Duplicate aliquots of the combined ethanolic extracts of each sample were used for 700 nm was provided by a mixture of super metal halide and incandescent lamps (Downs and Thomas, 1990). Plants were determination of free amino acid composition. Free amino acids in soluble N fractions were derivatized with phenylisothigrown with a 14-h photoperiod for 36 DAT (until V6 or V7) and then switched to a 12-h photoperiod for the remainder ocyanate and separated and quantified using reversed-phase NAKASATHIEN ET AL.: NITROGEN SUPPLY AND SEED PROTEIN CONCENTRATION IN SOYBEAN 1279 high performance liquid chromatography (Heinrikson and to maturity (control), (ii) 10 mM N from 4 DAT to V5 and 30 mM N from V5 to maturity, and (iii) 10 mM N from 4 DAT Meredith, 1984). to R5 and 30 mM N from R5 to maturity. These treatments were arranged in a randomized complete Total Seed Storage Protein Profiles block design with four replications (two replications in each Flowers were tagged on the day of opening, and then pods chamber). The statistical analysis revealed no significant were collected at 15, 30, 45, and 60 d after tagging. Seeds were chamber effect on any parameters; thus, a randomized comseparated from pod wall, frozen in liquid N, and stored in a plete block design with four replications is appropriate for freezer at 2808C until the analysis could be performed. Seeds this experiment. The data for all measured parameters were were extracted by homogenizing in 1:10 (w/v) 0.03 M Trisanalyzed using the ANOVA procedure of the Statistical AnalHCl buffer, pH 8.0, and 7.7 mM NaN3, using a Brinkman ysis System (Goodnight, 1982). If treatment effects were signifPolytron. Homogenates were centrifuged at 10 000 g using the icant at the 0.05 probability level, LSD0.05 values were calcuRC5C centrifuge (Sorvall Instruments, Dupont, Wilmington, lated for comparison of means. DE). The soluble protein concentration in the extracts was A randomized complete block design with three replicathen determined by the method of Bradford (1976) with botions was used for analysis of free amino acid composition vine plasma gamma globulin (Bio-Rad Inc., Richmond, CA) data. Treatments consisted of a 3 by 2 factorial with three as the standard. Seed proteins were denatured in 0.03 M Trissoybean varieties and two N levels. The data for the amino HCl buffer, pH 8.0, 69 mM SDS, and 0.28 M b mercaptoethaacid concentration and composition were analyzed using the nol in a boiling water bath for 10 min. ANOVA procedure of the Statistical Analysis System (GoodElectrophoresis was carried out using the buffer system night, 1982). If treatment effects were significant at the 0.05 described by Chua (1980) with 12.5 to 20% linear gradient probability level, LSD0.05 values were calculated for comparipolyacrylamide gel of 1.5-mm thickness for 18 h at 6 mA. Gels son of means. were stained with Coomassie brilliant blue R-250 as described For total seed storage protein profile measurements, two in Harlow and Lane (1988). Total seed protein profiles were replicates of treatment combinations as described above were quantified with a Molecular Dynamics Personal Densitometer used for 30, 45, and 60 DAF. Because the plants reached R5 SI equipped with a HeNe laser light source and ImageQuant stage after 15 DAF, a 3 by 2 factorial of three varieties and software for volume integration of total optical density of two N levels was applied at 15 DAF. Data are reported as entire protein bands (Molecular Dynamics, Sunnyvale, CA). the percentage of each individual storage protein band relative The amount of protein on gels was adjusted to ensure that to the total amount of storage protein on the gel at each the most intense bands were within the linear response range sampling date. of the detector. Storage protein subunits were identified by comparison with molecular weight standards run in outside RESULTS lanes on each gel. The relative amount of each storage protein subunit was expressed as a percentage of total storage protein Nitrogen Assimilation in the gel lane. Cultivars NC 107 and N87-984-16 had similar seed yields and whole plant dry weights, but both had signifiExperimental Design and Statistical Analysis cantly higher seed yields than NC 111 (Table 2). Relative For growth parameters and N analysis, the treatment structo the control, 30 mM N supplied from V5 and R5 to ture was defined based on two levels of N, 10 mM N (control) maturity increased seed yield of NC 107 and N 87-984-16 and 30 mM N applied during two different growth stages, V5 by an average of 45 and 55%, respectively, and decreased to maturity and R5 to maturity. Plant samples were harvested seed yield for NC 111 by 38 and 30%, respectively (Table at two different stages, R5 and maturity. This resulted in the following treatment combinations: (i) 10 mM N from 4 DAT 2). The 30 mM N treatments had no significant effect Table 2. Effect of 30 mM external N supplied after V5 and R5 on whole plant dry matter accumulation and seed yields. Values for individual treatments are means of four replicate plants. Time of N Whole plant Genotype application N concentration Seed yield dry matter