Rebetzke Et Al.: Factors Influencing Palmitate Content of Soybean Seed Oil

intake contributes to raised blood serum cholesterol levels, thus increasing the risk of coronary heart disease Dietary concerns over high saturates contained in edible vegetable (Willett, 1994; Uusitalo et al., 1996). Public perception oils has stimulated development of soybean [Glycine max (L.) Merr.] cultivars with reduced palmitate content. Little is known of factors that of this potential health issue has provoked a strong trend might influence phenotypic expression of palmitate content among towards greater consumption of foods containing lower soybean populations varying for presence of a major reduced palmilevels of saturated fats (Wilson, 1991; Uusitalo et al., tate allele. The objective of this study was to investigate how environ1996). In turn, the U.S. Food and Drug Authority has ment and genetic background influence palmitate content when introproposed labeling regulations indicating “low saturate” ducing the reduced palmitate trait into adapted backgrounds. Crosses vegetable oils must contain no greater than 70 g kg 1 were made between reduced palmitate germplasm, N87-2122-4 (53 g total saturates. Although soybean oil is relatively low kg 1 palmitate) and normal palmitate cultivars, A3733, Burlison, Kenin total saturates (≈120–200 g kg 1 ), at least a minimum wood, P9273, and P9341 (103–123 g kg 1 palmitate). For each cross, 50% reduction in saturated fat is needed to enhance F4:6 lines homozygous for major reduced or normal palmitate alleles the utility of soybean oil in this new market. were bulked separately into Maturity Groups (MG) II, III, IV, and V, and evaluated in 10 contrasting field environments during 1993. Palmitate is the predominant saturated fatty acid in Palmitate content varied between 82 and 90 g kg 1 across southern soybean and most other vegetable oils (Weiss, 1983). U.S. and Puerto Rican environments. Much of this environmental Major alleles conditioning reduced palmitate content variation was associated with changes in minimum temperature during are available in soybean germplasm obtained from rethe growing season. Genetic background effects were highly significant current selection (Burton et al., 1994) and chemical mu(P 0.01) with cross means for palmitate content ranging between tagenesis (Horejsi et al., 1994; Wilcox et al., 1994). In 81 and 93 g kg 1. Across different maturity groups, palmitate content response to consumer demand for healthful oils, soyof the progeny was correlated (r 0.94–0.99, P 0.05) with mean bean breeders are now actively incorporating the recontent of the normal palmitate parent, such that for every 1 g kg 1 duced palmitate trait into cultivar development propalmitate increase in the normal palmitate parent there was a 0.32 grams (Burton et al., 1996). Given the need to transfer to 0.51 g kg 1 palmitate increase in the progeny. Genetic background effects were presumed to be associated with action of minor alleles reduced palmitate alleles into a range of adapted, hightransmitted from the normal palmitate parent. Presence of the reduced yielding genetic backgrounds (Wilson, 1991; Burton et palmitate allele was associated with significantly (P 0.01) lower al., 1996), it would be helpful to understand whether stearate ( 6 to 13%) and higher oleate ( 4 to 10%) contents palmitate expression is contingent upon reduced palmiacross all maturity groups. Selection of low palmitate, high-yielding tate genes contributed by the gene donor only, or genes parents should further decrease palmitate content and produce correpresent within both donor and adapted parents. The lated improvements in stearate and oleate contents to improve overall interaction of target genes with genes present within oil quality in progeny containing reduced palmitate alleles. recipient genetic backgrounds can be an important consideration when introducing new traits into a commercial breeding program (Hallauer and Miranda, 1988). E evidence and clinical studies alike Further, it would be useful to document the influence have demonstrated that higher saturated fatty acid of genetic background on expression of the reduced palmitate trait in different maturity zones and soybean G.J. Rebetzke, CSIRO Plant Industry, P.O. Box 1600 Canberra ACT production regions of the USA. 2601 Australia; V.R. Pantalone, Dep. Plant and Soil Science, UniverThe influence of parental selection and genetic backsity of Tennessee, Knoxville, TN 37901; J.W. Burton, T.E. Carter, Jr., ground on phenotypic expression for reduced palmitate and R.F. Wilson, USDA-ARS, and Crop Science Dep., North Carolina State Univ., Raleigh, NC 27695. Cooperative investigations content has not been reported for soybean. Earlier studof the USDA-ARS, and North Carolina Agric. Res. Serv., Raleigh, ies (Horejsi et al., 1994; Rebetzke et al., 1998) showed NC. Received 22 May 2000. *Corresponding author (G.Rebetzke@ that palmitate content measured in different soybean pi.csiro.au). Abbreviations: MG, Maturity Group. Published in Crop Sci. 41:1731–1736 (2001). 1732 CROP SCIENCE, VOL. 41, NOVEMBER–DECEMBER 2001 15 lines. The inclusion of several random lines produced bulks populations reflected variation at both major and minor that were putatively near-isogenic for maturity and major palloci. The objective of this study was to evaluate the mitate alleles but were random for minor palmitate alleles influence of environment and genetic background on and alleles conditioning other traits. palmitate content in soybean populations developed The eight bulks per cross representing four maturity groups from crosses between reduced palmitate germplasm and (MG I, II, IV, and V) and contrasting palmitate level were normal palmitate cultivars. evaluated in 1993 along with parental (MG II–III) and nonparental (MG IV–V) check lines. Studies were conducted MATERIALS AND METHODS at three midwestern (Johnson City, IA; Napoleon, OH: St. Joseph, IL) and seven southern (Union City, TN; early and Reduced palmitate germplasm, N87-2122-4 (53 g kg 1 pallate sowing at Clayton, NC; Fletcher, NC; Plymouth, NC; mitate), was used as a female in crosses to randomly selected Greenville, MS; Isabella, Puerto Rico) environments to promidwestern cultivars, A3733, Burlison, Kenwood, P9273, and vide a broad range of conditions during the growing seaP9341 to generate five segregating populations. N87-2122-4 is son (Table 1). The MG II and III bulks were sown in the of MG V and traces its pedigree to the reduced palmitate Midwest and at Union City, TN, while the MG IV and V germplasm release, N79-2077-12 (Burton et al., 1994). The bulks were sown in all seven southern environments. The parental cultivars ranged in maturity from MG II to III and experimental design at each location was a randomized comrepresented some of the highest-yielding midwestern cultivars plete block with three replicates. Three-row plots, 4.8 m in at the time of the study. The F1 plants were grown in Puerto length and with 0.38 m between rows, were used at all locations Rico and harvested F2 seed sown at Clayton, NC, in 1991. A except Fletcher (2.4-m-long plots); Greenville, Union City, 5-g seed sample was obtained from individually harvested F2 and the Midwest (four-row plots, 6.1 m long). Seed was harplants and analyzed for fatty acid composition following Wilvested at maturity from the center row(s) at each location. son et al. (1981). Seed containing either major alleles for Maturity was determined for all plots as the date at which reduced or normal palmitate can be genotyped from the quan95% of pods in the row obtained mature color. Seed from tity of palmitate found in their oil (Wilcox et al., 1994; Reeach plot was analyzed for oil composition following Rebetzke betzke et al., 1998). Eighty reduced and 80 normal palmitate et al. (1998) and for seed oil content using near-infrared analyprogeny from each cross were selected and inbred through ses by the USDA National Center for Agricultural Utilization single-seed descent to the F4:5 generation. The reduced and Research (NCAUR), in Peoria, IL. Mean daily minimum and normal palmitate classes were presumed to differ for presence maximum temperatures were obtained for the growing season or absence of the reduced palmitate allele described by Wilcox at all sites (Table 1). et al. (1994) and Rebetzke et al. (1998). A combined analysis of variance over environments was In 1992, a single row of each F4:5 line was grown at Clayton, conducted on each maturity group for seed oil traits using the NC, with MG I to VI check cultivars for the purpose of matuSAS procedure GLM (SAS Institute, 1990). Error variances rity designation. Seed was harvested and the putative reduced for each environment were deemed homogenous following or normal palmitate genotype verified for each line by fatty the Fmax test for homogeneity of error variances (Sokal and acid analysis of seed oil following Rebetzke et al. (1998). Equal Rohlf, 1981). Palmitate content (reduced vs. normal bulks) numbers of seed were then bulked for each cross from F4:6 was deemed a fixed effect, and crosses and environments random lines of similar maturity (maximum maturity range in each effects in deriving appropriate errors for statistical testing. bulk was 5 d) and either reduced or normal for palmitate content. Individual bulks were composites of between 12 and Protected least significant differences (LSD) were calculated Table 1. Environment means for seed oil quality characteristics measured in 1993 on soybean bulk populations representing soybean Maturity Groups (MG) II to V. Fatty acid content Climate† Temperature Oil Environment Palmitate Stearate Oleate Linoleate Linolenate content Rainfall Min. Max.