Crop Science Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 24 February 2006
Published in Crop Sci 46:747-750 (2006)
© 2006 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
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 (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Beuselinck, P. R.
Right arrow Articles by Bilyeu, K. D.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Beuselinck, P. R.
Right arrow Articles by Bilyeu, K. D.
Agricola
Right arrow Articles by Beuselinck, P. R.
Right arrow Articles by Bilyeu, K. D.
Related Collections
Right arrow Crop Genetics
Right arrow Soybean

CROP BREEDING, GENETICS & CYTOLOGY-NOTE

An Assessment of Phenotype Selection for Linolenic Acid Using Genetic Markers

P. R. Beuselincka,*, D. A. Sleperb and K. D. Bilyeua

a USDA-ARS, Plant Genetics Research Unit, Univ. of Missouri, Columbia, MO 65211
b Dep. of Agronomy, Univ. of Missouri, Columbia, MO 65211

* Corresponding author (BeuselinckP{at}missouri.edu)


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Marker assisted selection (MAS) procedures potentially can make breeding more efficient, as genotypes can be identified before pollination allowing breeders to cross or backcross only suitable materials. Our objective was to assess the accuracy of our phenotypic selections in soybean [Glycine max (L.) Merr.] for linolenic acid using molecular markers specific for mutations in two fatty acid desaturase genes, GmFAD3A and GmFAD3C. The markers were not available earlier in the selection process and were used retrospectively to determine that phenotypic selection for seed linolenic acid ≤ 35 g kg–1oil was successful in capturing genotypes homozygous for mutant alleles at both loci, but phenotypic selection was not perfect. Chemical analysis for seed linolenic acid concentration was not an accurate predictor of genotype. The advancement of heterozygotes reduced the selection efficiency relative to what would have been possible using molecular markers specific for mutations in the two fatty acid desaturase genes. Errors are thought to have derived from inaccurate sample tracking or identification, contamination, or errors in chemical analyses. Use of mutation-specific molecular markers to identify F2 lines homozygous for mutant alleles in GmFAD3A and GmFAD3C, combined with diligence in reducing sampling errors, would eliminate the need for chemical testing for linolenic acid content in subsequent generations where screening can emphasize other traits.

Abbreviations: MAS, marker assisted selection • MG, maturity group • PCR, polymerase chain reaction


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
A DIFFICULT PROBLEM for the plant breeder has been to discern to what extent a trait is heritable and the result of gene action, and to what extent the trait is the result of an environmental influence. The distinction between what is a genetic vs. an environmental effect for quantitative traits is generally difficult to discern. The ideal situation for selection of a quantitative trait is that the trait has high heritability and the phenotype can be observed in all individuals before reproductive age (Dekkers and Hospital, 2002). Unfortunately, the phenotype for most quantitative traits is an imperfect predictor of the breeding value of an individual. Until recently, breeders' selections on economically important quantitative traits have been based on phenotype, without information on the genetic architecture corresponding to the traits. Knowing the number, positions, and effects of all the genes involved could make trait improvement efficient.

Linkages established between nonfunctional or anonymous markers with quantitative trait loci (QTL) are fairly common, but they are subject to losing their association with the desired phenotype through recombination. Molecular markers for causal mutations for quantitative traits are harder to find than simply inherited traits because of the extensive analysis required for positional cloning or the availability of suitable candidate genes. However, these perfect molecular markers lack the ambiguity associated with QTL-type markers, and thus eliminate the problem of disassociation of the trait from a linked anonymous locus. Marker assisted selection allows selection at the genotype level, or specifically the desired alleles, one time, without the need to track and stabilize phenotypes across several generations. Plants with the desired genotype can be identified before pollination with MAS, allowing breeders to more efficiently cross or backcross. Mutation-specific molecular markers can also be used to validate the genetic constituency of phenotype-based selections made on the assumption of a corresponding desirable genotype.

Omega-3 fatty-acid desaturase genes (FAD3) control seed linolenic acid levels in soybean. Previously in our lab we used database homology searches and gene cloning to identify and characterize three soybean microsomal omega-3 fatty-acid desaturase genes that contribute to seed linolenic acid levels (Bilyeu et al., 2003). The complete coding sequences for three soybean microsomal FAD3 genes have been assembled and are available in GenBank (Benson et al., 2002). We screened the soybean FAD3 homologs for mutations, and developed molecular markers for the low linolenic acid soybean line CX1512–44 that contains defects in the GmFAD3A and GmFAD3C genes (Bilyeu et al., 2003, 2005).

Having a set of molecular markers for the GmFAD3A and GmFAD3C genes afforded us a unique opportunity to conduct a retrospective assessment of the effectiveness of breeding selections for low linolenic acid based on the phenotype of this quantitative trait. Our objective was to assess the accuracy of our phenotypic selections in soybean for linolenic acid using molecular markers specific for mutations in two fatty acid desaturase genes, GmFAD3A and GmFAD3C.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soybean Lines
The low linolenic acid soybean line CX1512–44 was previously developed from a mutagenesis breeding program as a selection from the cross 10689–4 x C1813 (J.R. Wilcox, 2004, personal communication). The 10689–4 parent is a low linolenic acid mutant line. C1813 is a selection from the cross C1655 x ‘Pella 86’. The parentage of C1665 is ‘Nebsoy’ (Williams et al., 1980) x A75–305022, and A75–305022 is a selection from the cross ‘Wye’ x [‘Amsoy’ (Weber, 1966) x ‘Wayne’ (Bernard, 1966)].

The line CX1512–44 [Maturity Group (MG) 3.0] and two experimental breeding lines, SS00–2924 (MG 3.0) and SS00–2617 (MG 3.0), were planted, as part of a larger breeding effort, at the Bradford Research and Extension Center, located near Columbia, MO, in 2001. SS00–2924 and SS00–2617 were reciprocally hand-crossed with CX1512–44, and resulting F1 seed were bulked within crosses to produce two populations. Population 224 was derived from SS00–2924 x CX1512–44, and Population 233 was derived from SS00–2617 x CX1512–44. F1 plants from the 224 and 233 populations were grown in Guacima, Costa Rica, in winter 2001–2002. In January 2002, 150 F2 seeds were chipped from each population, then chips were express shipped to Columbia, MO, and analyzed for linolenic acid concentration. Of the 300 chipped seed, 10 were chosen for a desirable linolenic acid content of ≤ 35 g kg–1 oil from Population 224, and seven were from Population 233. The 17 chipped F2 seeds were planted in rows with 10 cm between plants and 74 cm between rows in Nazareth, Costa Rica, in February 2002; single-plant threshed in May 2002; and all F3 seed were planted in Columbia, MO, in June 2002. Additional selection for yield and resistance to soybean cyst nematode (Heterodera glycines Ichinohe) advanced 65 selected F3's for further development. Sixty-three F3's were selected from Population 224, and two were from Population 233. The 65 F3's were single-plant threshed and three whole F4 seeds from each plant were individually analyzed for linolenic acid to provide a mean linolenic acid level for each F3. Approximately 100 F4 progeny from each F3 were planted in late December 2002 into separate 3-m progeny rows with 10 cm between plants and 74 cm between rows in Upala, Costa Rica, before obtaining results from linolenic acid determinations on the F3:4 seed. In April 2003 leaf squashes were collected on FTA (Whatman, Clifton, NJ) cards from four randomly chosen F4 plants from each of the 65 progeny rows. The four plants in each row were genotyped independently. The FTA cards were express shipped to Columbia, MO, for DNA analysis to assign a genotype to the F3 parent of the four sampled F4 plants.

Phenotype Analysis
For fatty acid determination, chips were taken distal to the embryonic axis ({approx}20% of the seed) from 150 F1 derived F2 seed for Populations 224 and 233, or three whole seeds from each of the 65 F3 plants were analyzed. Linolenic acid in each seed sample was determined as a proportion of total fatty acids and represented as g kg–1 of oil by the FAME gas chromatography method. Chips from seeds and individual whole seeds were analyzed as separate samples. Crushed samples were extracted overnight in 0.5 mL of chloroform–hexane–methanol (8:5:2, v/v/v) for chips or 1 mL for single seeds. Derivatization of 0.5 mL of chip solvent or 150 µL of single seed solvent was done with 75 µL of methylating reagent (0.5 M methanolic sodium methoxide–petroleum ether–ethyl ether, 1:5:2, v/v/v). For single seeds, samples were diluted with hexane to {approx}1 mL. An Agilent (Palo Alto, CA) series 6890 capillary gas chromatography fitted with a flame ionization detector (275°C) was used with an AT-Silar capillary column (Alltech Associates, Deerfield, IL). Standard fatty acid mixtures (Animal and Vegetable Oil Reference Mixture 6, AOACS) were used as controls.

Genotype Analysis
Molecular markers for mutant alleles of GmFAD3A and GmFAD3C loci were developed using a detection method for single nucleotide polymorphisms based on the McSNP assay (Ye et al., 2002). The genomic region of interest was first amplified by polymerase chain reaction (PCR) followed by a restriction enzyme digestion which distinguishes between wild-type and mutant alleles. We used a real-time PCR instrument (MJ Research Opticon, Waltham, MA) and the double-stranded DNA binding dye SYBR Green I (Molecular Probes, Eugene, OR) for a melting curve analysis so that PCR and digestion products could be characterized without separation on agarose gels. For GmFAD3A, the primers were 3AD1 (TTGCATCACCATGGTCATCAT) and 3AIX (AGCTATTATCTAGCATTAACCTCA). For GmFAD3C, the primers were 653Dup (GTCCTTTGTTGAACAGCATT) and 653T (CTCCTGCAAAAAATCCATGAGTTGT). The PCR templates consisted of 2-mm washed FTA (Whatman, Clifton, NJ) card punches prepared from leaves according to the manufacturer's instructions. The 15-µL reactions for PCR contained template, buffer [40 mM tricine-KOH (pH 8.0), 16 mM KCl, 3.5 mM MgCl2, 3.75 µg mL–1 BSA, 200 µM dNTPs], 10% DMSO, 0.5 µM each primer, 0.25X SYBR Green I, and 0.2X Titanium Taq polymerase (BD Biosciences, Palo Alto, CA). Amplification conditions were 95°C for 5 min, 35 cycles of 95°C for 20 sec, 60°C for 20 sec, and 72°C for 20 sec. Restriction enzyme reactions were performed for 14 to 18 h in the same tube after the addition of 30 µL of a mix containing 20 µL 2x MaeIII buffer (Roche Applied Science, Indianapolis, IN), 9.85 µL ddH2O, and 0.15 µL MaeIII (1.67 U µL–1, Roche Applied Science, Indianapolis, IN) for GmFAD3A or 3.5 µL 10x buffer 2 (New England Biolabs, Beverly, MA), 26.3 µL ddH2O, and 0.2 µL NcoI (10U µL–1, New England Biolabs, Beverly, MA) for GmFAD3C. GmFAD3A MaeIII reactions were incubated at 55°C while GmFAD3C NcoI reactions were incubated at 37°C. Melting curve analysis followed restriction enzyme digestion of PCR products with parameters of 70 to 90°C with 0.2°C increases and reads every 1 s. For GmFAD3A, wild-type alleles (100- and 48-bp products) produced a peak at 78°C, mutant alleles produced a peak at 82.5°C (148 bp) and heterozygotes produced both peaks (148, 100, and 48 bp). For GmFAD3C, wild-type alleles (134- and 59-bp products) produced a peak at 79.5°C, mutant alleles produced a peak at 81.5°C (193 bp), and heterozygotes produced both peaks (193, 134, and 59 bp). In some cases, products were resolved on 1.5% agarose gels. Only when all four F4 samples contained homozygous alleles was the F3 parent assigned a homozygous genotype for either GmFAD3A or GmFAD3C. Genotype identifications (CX1512-44 and F3's) were compared with linolenic acid profiles obtained from GC analysis to determine efficiency of chemical phenotype selection vs. MAS.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
The relatively high linolenic acid content in the fatty-acid profile of most current soybean cultivars is responsible for oxidative instability and undesirable flavors in the oil (Dutton et al., 1951). Although progress has been made using induced mutation followed by phenotypic selection to achieve significantly low linolenic acid levels, selection has been hampered by the quantitative nature of the trait and interactions with environment. At least three independent genetic loci (fan, fan2, fan3) are associated with seed linolenic acid levels, with mutant alleles identified at all loci including fanx (Fehr et al., 1992; Rahman and Takagi, 1997; Ross et al., 2000; Wilcox and Cavins, 1987). Multiple alleles at fan have been reported and the locus has been mapped (Brummer et al., 1995; Rahman et al., 1996; Rennie and Tanner, 1991; Stojsin et al., 1998; Wilcox and Cavins, 1987). We previously identified GmFAD3A as the fan locus (Bilyeu et al., 2003). GmFAD3C potentially represents fan2 or fan3 (W.R. Fehr, 2005, personal communication). In our laboratory we developed molecular markers for mutant alleles of GmFAD3A and GmFAD3C genes derived from CX1512–44 to enable screening for low linolenic acid soybean lines (Bilyeu et al., 2003, 2005). For those experiments, there was complete correspondence between the low linolenic acid phenotype for seeds that developed under defined growth chamber conditions and the homozygous mutant genotype. Other mutants for low linolenic acid trait would require distinct molecular markers (data not shown). In this work, our molecular marker analysis allowed us to assign a genotype to the F3 parent. It was assumed that the effect of the field growth environment might alter the linolenic acid content of the seed, but the molecular markers would provide a perfect predictor of phenotypic segregation in subsequent generations. A retrospective evaluation of the genotype of the F3 parent allowed us to determine phenotypic selection efficiency for linolenic acid as typically done by the breeder at the F2 and F4 generations.

The criterion for selecting lines that were chipped and analyzed as F2 seed was ≤35 g linolenic acid kg–1 oil. Selection in the F3 generation among the linolenic acid lines for yield and resistance to soybean cyst nematode reduced the number of lines chosen for further development to 65, and these lines descended from 17 F2 seeds. After individually threshing single F3 plants, three F4 seeds were analyzed for fatty acid content while the remaining seeds were planted as F4 progeny rows. We sampled F4 individuals within progeny rows to assign an F3 genotype (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Seed linolenic acid of 57 F3 genotypes with wild-type or mutant alleles for two omega-3 fatty acid desaturase genes, GmFAD3A and GmFAD3C.

 
Five of nine possible genotypes were assigned to the F3 parents for 57 of the 65 F4 progeny rows; the remaining eight lines yielded inconclusive identification of their genotype. The homozygous genotype for both mutant genes (aacc) was found most often and corresponded to a linolenic acid phenotype of F4 seeds most similar to the heterozygous genotype aaCc. It is apparent that linolenic acid concentrations of the F4 seed were not an accurate predictor of genotype. It is notable that eight lines with at least one wild-type GmFAD3A allele were present in the F3:4 populations. The advancement of these lines may indicate an earlier sample tracking or identification error, contamination, or analytical error since the presence of even one wild-type GmFAD3A allele should have raised the linolenic acid levels above the F2 threshold of 35 g kg–1 oil (Bilyeu et al., 2005). It cannot be determined in this study if the range of values in linolenic acid concentration for each of the five genotypes recovered reflect an undefined environmental growth effect or analytical error. Without a complete set of the nine possible genotypes it was impossible to assign a substitution value for each allelic combination of the two mutations. In other experiments, we have demonstrated that each wild-type allele of the GmFAD3C gene can contribute to the linolenic acid level by {approx}4 to 11 g kg–1 oil (Bilyeu et al., 2005).

Similar to the F2 generation phenotyping, the limit established for selecting a low linolenic acid F4 line was a mean of 35 g kg–1 oil; that is, a null hypothesis (H0) was intuitively established that the chemical phenotype (i.e., linolenic acid content ≤ 35 g kg–1 oil) of a selected F4 line was not determined by the homozygous mutant genotype, aacc; that is, the alternate hypothesis (HA) was that the chemical phenotype of a selected F4 line was determined by the homozygous mutant genotype, aacc. For the phenotypic selection, the linolenic acid content for 48 of the 57 genotyped F4 progeny met the selection criterion.

The genotyping assays revealed the efficiency of phenotypic selection. Nine F4 progeny included among the phenotypic selections for linolenic acid concentration ≤ 35 g kg–1 oil were heterozygous (8 aaCc and 1 AaCC) and one was homozygous mutant for only one locus (aaCC); for these genotypes, the H0 was rejected and the HA accepted. Accepting the HA constituted a Type-I error of {approx}20%. This type of error would add additional uncertainty to anticipated outcomes from phenotypic selections made for low linolenic acid and necessitate additional phenotypic selections to ensure the stable inheritance of the trait. One F4 progeny was homozygous mutant for both genes, but had a mean linolenic acid concentration of >35 g kg–1 oil and would not have been chosen using the selection criterion. By accepting the H0, a Type-II error (Steele and Torrie, 1960) was made, but the error had little consequence in the breeding scheme because other genotypes were selected without error.

From among the progenies of selections with a linolenic acid concentration ≤ 35 g kg–1 oil, reselections were made for overall acceptable yield and resistance to soybean cyst nematode; 16 lines, all from Population 224, were advanced for yield testing in 2004. Of the 16 lines, 13 were homozygous mutant (aacc) and three were heterozygous (aaCc). The inclusion of the heterozygotes constituted a Type-I error of {approx}19%.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Molecular markers for GmFAD3A and GmFAD3C were developed after the initial crosses and phenotypic selections were made, so the markers were not available earlier in the selection process. We used these markers retrospectively to determine that phenotypic selection for seed linolenic acid ≤ 35 g kg–1 oil was successful in capturing homozygous mutant genotypes, but phenotypic selection was not perfect. We were alerted that our linolenic acid concentration analysis of the seed was not an accurate predictor of genotype. Mutation-specific markers, like those for GmFAD3A and GmFAD3C, can be used to identify genotypes of parental lines before crossing, confirm F1's, and select F2's. Current work focuses on backcrossing mutant alleles that provide the greatest reduction in seed linolenic acid levels for soybean breeding programs.


   ACKNOWLEDGMENTS
 
We are grateful for the advice and consultation provided by statisticians Drs. Richard Madsen (retired) and Mark Ellersieck of the University of Missouri, Columbia, MO.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Contribution of the Missouri Agric. Exp. Stn. Mention of a trademark, vendor, or proprietary product does not constitute a guarantee or warranty of the product by the USDA or the Univ. of Missouri and does not imply its approval to the exclusion of other products or vendors that may also be suitable.

Received for publication September 2, 2005.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Crop Sci.Home page
W. R. Fehr
Breeding for Modified Fatty Acid Composition in Soybean
Crop Sci., December 18, 2007; 47(Supplement_3): S-72 - S-87.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
G. E. Byfield and R. G. Upchurch
Effect of Temperature on Microsomal Omega-3 Linoleate Desaturase Gene Expression and Linolenic Acid Content in Developing Soybean Seeds
Crop Sci., November 7, 2007; 47(6): 2445 - 2452.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
K. Bilyeu, L. Palavalli, D. A. Sleper, and P. Beuselinck
Molecular Genetic Resources for Development of 1% Linolenic Acid Soybeans
Crop Sci., July 25, 2006; 46(5): 1913 - 1918.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
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 (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Beuselinck, P. R.
Right arrow Articles by Bilyeu, K. D.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Beuselinck, P. R.
Right arrow Articles by Bilyeu, K. D.
Agricola
Right arrow Articles by Beuselinck, P. R.
Right arrow Articles by Bilyeu, K. D.
Related Collections
Right arrow Crop Genetics
Right arrow Soybean

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