| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Crop Sciences, University of Illinois, Urbana, IL 61801
b Plant Science Dep., South Dakota State University, Brookings, SD 57007
c Chromatin, Inc., 60 Hazelwood Dr., Champaign, IL 61820
d Dep. of Agronomy and Horticulture, University of Nebraska, Lincoln, NE 68583
* Corresponding author (bdiers{at}uiuc.edu)
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: AFLP, amplified fragment length polymorphism • BC, backcross • BSA, bulked segregant analysis • cM, centimorgan • LG, linkage group • MG, maturity group • PI, plant introduction • PCR, polymerase chain reaction • QTL, quantitative trait loci • RFLP, restriction fragment length polymorphism • SSR, simple sequence repeat
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Diers et al. (1992) used genetic markers to study the genetic control of seed protein and oil concentration. They detected two major QTL controlling protein concentration in a population developed from a cross between PI 468916, an unadapted G. soja accession with high protein concentration that was collected at Liaoning, China, and A81–356022, an adapted maturity group III G. max breeding line. Lines homozygous for the G. soja allele at the most significant molecular marker for each QTL were associated with an increase in protein of 24 g kg–1 for LG K (now known as LG I) and 17 g kg–1 for LG A (now known as LG E) when compared with lines homozygous for the G. max allele.
Sebolt et al. (2000) continued this research by backcrossing the high-protein alleles on LG I and LG E from the G. soja donor parent into A81–356022, the adapted recurrent parent. A backcross-three (BC3) population was developed and tested for agronomic performance, seed protein, and oil concentration, and genotypes on LG I and LG E were determined by means of SSR markers. The authors confirmed the G. soja allelic effect of the LG I protein QTL but detected no significant effect in the region where the LG E QTL mapped. For the LG I QTL, lines homozygous for the PI 468916 marker alleles had average protein concentrations that were significantly (P < 0.001) greater (21 g kg–1) than lines homozygous for the G. max alleles. Furthermore, the authors reported that lines homozygous for the G. soja marker alleles had significantly greater plant height, reduced yield and oil concentration, reduced seed weight, and earlier maturity than lines homozygous for G. max alleles.
Sebolt et al. (2000) created three additional populations to study the effect of the LG I protein QTL in different genetic backgrounds. These populations were developed by crossing the cultivars Kenwood and Parker and the high protein line C1914 to a line from the BC3 population that carried the high protein allele for the LG I QTL. The effect of the high protein allele was significant in the Kenwood and Parker populations but not in the C1914 population. These results led Sebolt et al. (2000) to speculate that C1914 had a high protein allele on LG I that was allelic with the G. soja allele.
Chung et al. (2003) reported the mapping of a protein QTL allele to LG I from PI 437088A, a G. max accession with a high protein concentration. The QTL mapped to the same region on LG I as the QTL reported by Sebolt et al. (2000). They observed an 18 g kg–1 increase in protein concentration among lines homozygous for the allele from PI 437088A compared with lines homozygous for the allele from the elite parent. Similar to the QTL found by Sebolt et al. (2000), the high protein QTL detected by Chung et al. (2003) was associated with lower oil concentration, reduced yield, and earlier maturity.
Results from these studies show that plant introductions can be successfully used as sources for increasing protein concentration in soybean. What remains unclear, however, is whether reduced oil concentration and lower yield associated with increased protein concentrations arise from pleiotropic effects of the LG I protein QTL allele from G. soja, or simply arise from coupling-phased G. soja alleles at loci governing oil and yield that are tightly linked to the protein QTL.
The first objective of this study was to fine map the protein QTL on LG I using populations that either segregated or were fixed in the region where the QTL has been mapped. The second objective was to determine whether inverse relationships between protein concentration and yield and oil concentration may be diminished through chromosomal recombination near this QTL.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The location of the protein QTL was further narrowed in the second set (Set 2), which included four BC5 populations. These populations were developed by first making a fifth backcross onto lines from P-85 in Set 1. The BC5 populations were inbred to the BC5F5 generation, and during each generation of inbreeding, only plants that were heterozygous for the region where the LG I protein QTL maps were selected. Approximately 1500 BC5F4 plants and 1000 BC5F5 plants were then tested with the SSR markers Satt127, Satt700, Satt239, and Satt496, which map close to the LG I protein QTL, to identify plants with genetic recombination near to the QTL. C602–15–5, C602–15–6, and C609–45–2 were BC5F4 individuals selected based on their heterozygosity only within the QTL region due to recombination. Populations of BC5F5–derived lines were formed from the selected plants and P-C602–15–5 contained 69 lines, and both P-C602–15–6 and P-C609–45–2 contained 53 lines. P-C602–15–5 and P-C602–15–6 were derived from different BC5F4 plants, but they were descended from a single BC5F3 plant. P-C609–45–2-2 was developed to confirm the results from P-C609–45–2 and originated from a single BC5F5 plant from P-C609–45–2 that was heterozygous for Satt496. P-609–45–2-2 contained 39 BC5F6–derived lines.
Agronomic Performance Evaluation
Set 1
The Set 1 populations were evaluated in field tests near Urbana, IL, from 1999 to 2001. Each population was grown in a separate block, and plots were arranged within blocks in a randomized complete block design with two replicates each year. BC4F3:5 seed were planted on 7 May 1999, BC4F3:6 seed were planted on 3 May 2000, and BC4F4:7 seed were planted on 2 May 2001. The adapted parent, A81–356022, and the cultivar Macon were included as checks within each block. Two-row plots with a 3.60 m length and a 0.76 m row spacing were planted at a rate of 34 seed m–1. Field plots were rated each year for plant height, lodging, and date of maturity. Maturity was expressed as the number of days after 31 August when 95% of the plants within a plot had reached their mature pod color (R8) (Fehr et al., 1971). Plant height and lodging data were collected after plots were fully mature. Plant height was the mean distance from the ground to the terminal nodes of plants within plots. Lodging scores were collected using a visual scale ranging from 1 to 5, with 1 designating all plants erect and 5 designating all plants prostrate.
Plots were harvested with a plot combine at maturity to determine seed yield. Seed yields were computed on a 130 g kg–1 moisture basis. Seed size was determined only in 2000 and 2001 by weighing a 100 seed sample from each plot. Seed protein and oil concentration were measured from near-infrared transmittance of a 25-g seed sample taken from each plot. The protein and oil analyses were conducted at the USDA Northern Regional Research Center at Peoria, IL, and were reported on a moisture-free basis.
Set 2
Agronomic evaluations of the Set 2 populations were conducted during 2004 near Urbana, IL. The tests were planted on 17 June 2004. The lines were grown in a completely randomized design in one row plots with a 1.47-m length and a 0.76-m-row spacing and a seeding rate of 34 seed m–1. P-C609–45–2 was grown in three replications while all other populations were not replicated because of a lack of available seed. Data for agronomic performance and seed protein and oil content were collected as described for Set 1.
Molecular Marker Evaluation
DNA was extracted from leaves of at least 10 plants per line according to Kabelka et al. (2006). Polymerase chain reaction (PCR) procedures were performed according to conditions described by Cregan and Quigley (1997). The PCR products were analyzed by electrophoresis in 3% (w/v) Metaphor (FMC BioProducts, Rockland, ME) agarose gels, or in 6% (w/v) nondenaturing polyacrylamide gels and stained with ethidium bromide (Wang et al., 2003).
DNA samples of the lines in the Set 1 populations were assayed with SSR markers mapping near the LG I protein QTL. These markers were developed by Dr. P.B. Cregan (USDA-ARS, Beltsville, MD). A subset of 10 lines from each Set 2 population was assayed with these LG I markers to identify the marker-flanked segment near the LG I QTL segregating in each population. All lines in the Set 2 populations were evaluated with the LG I marker Satt496, which had been found to be one of the SSR markers closest to the protein QTL based on Set 1 results and Chung et al. (2003).
A bulked segregant analysis (BSA) (Michelmore et al., 1991) was completed to identify AFLP markers near the LG I protein QTL to further saturate the region with molecular markers and increase mapping resolution. P-85 and P-142 were both used in the BSA. For each population, DNA samples from five lines homozygous for A81–356022 alleles throughout the LG I segregating region were pooled into one bulk and DNA samples from five lines homozygous for PI 468916 alleles were pooled into a second bulk. The bulks were then screened with more than 500 combinations of AFLP primers to identify primer pairs that produced a fragment in one bulk but not in the other. All polymorphic AFLP markers were then tested on the entire population in which they were identified (i.e., either P-85 or P-142) and mapped relative to other markers in the region. The AFLP markers mapping to the LG I protein QTL region were tested on a subset of 10 lines from each population to further define the regions segregating in each population and help orient closely linked markers. The AFLP assays were conducted according to the protocol of Vos et al. (1995) with the modifications described by Kabelka et al. (2005).
Data Analysis
Set 1
The SSR and AFLP marker segregation data across all Set 1 populations were used to construct a genetic map of LG I within and across these populations on the basis of the Kosambi mapping function in JoinMap 3.0 (Van Ooijen and Voorrips, 2001).
Agronomic performance data for all entries except the checks were analyzed as a multiyear dataset for each agronomic trait with SAS PROC GLM (SAS Institute, Cary, NC). All factors were analyzed as random effects, and no combined analysis across populations was done because each population was grown in a separate test. Line and population means were then calculated for the agronomic traits, both within and over years, with the SAS GLM procedure. Single-factor analysis of variance with PROC GLM of SAS was used to test for associations between agronomic and seed traits and the allelic segregation at each marker. These analyses were done separately for each population.
Set 2
Single-factor analysis of variance was done with PROC GLM to test for associations between the segregation of Satt496 marker alleles and agronomic and seed traits for each population separately and in an analysis combined over populations. This combined analysis was done for the Set 2 populations because lines from all four populations were randomized together in a single experiment. In the combined population analysis, the significance of marker genotypes, populations, and the marker genotype x population interaction were tested.
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Results showing which markers were segregating in each NIL population were useful in orienting the markers. For example, although no recombinants were observed between Satt239 and Satt496 in our mapping populations, these markers were ordered on the basis of data from P-C609–45–2 and P-C609–45–2-2. In both populations, Satt239 and other markers completely linked to it or above it were not segregating, whereas Satt496 and ACG9a were segregating. Because the recombination between ACG9b and Satt496 had been selected previously, the most likely position of Satt496 is below Satt239, since this would require only one new recombination event, whereas mapping Satt496 above Satt239 would have required two recombination events.
Set 1 Evaluations
When data from each Set 1 population were analyzed across years, the line effect was significant in all combinations with the exception of yield in P-85 and lodging in P-142. The genotype x year interaction was significant for yield, lodging and height for P-85, yield and lodging for P-92, and lodging and maturity for P-142.
Data were combined across years for each population to test for significant marker associations. Markers on LG I were significantly (P < 0.001) associated with protein and oil concentration and maturity in each of the three populations (Fig. 1A) and were associated with seed yield and seed size in two out of three populations. Yield was significant for P-85 and P-142 but not P-92, whereas seed size was significant in P-92 and P-142 but not P-85. Lodging was not significant in any population, and height was significant (P < 0.05) only in P-92. These data show that lines homozygous for the introgressed G. soja marker alleles had greater seed protein concentrations, lower oil concentrations, earlier maturities, generally lower yields, and smaller seeds, and in one population, greater heights than did sister lines homozygous for the G. max marker alleles. For each population, the markers with the most significant effect for any given trait were Satt367 and Satt239, and the interval between them also had high LOD scores. The combined results across these Set 1 populations suggest that the QTL controlling these traits resides in an 11-cM interval between Satt614 and Satt354.
Set 2 Evaluations
The analysis across Set 2 populations revealed significant marker genotype and population effects. However, no significant marker genotype x population interaction effect was observed for any of the traits tested, so we did not test for differences in the size of the QTL effect among populations.
Analysis of the agronomic and seed data showed that the segregating region in all four populations had a significant effect on both protein and oil concentration (Fig. 1B). The analysis also showed significant effects on maturity in P-C602–15–6, P-C609–45–2, and P-C609–45–2-2; on lodging in P-C609–45–2; and on yield in P-C602–15–5 and P-C602–15–6. No association between marker genotypes and plant height was observed in any of the Set 2 populations. In the populations showing significant effects, lines homozygous for the QTL allele from PI 468916 had higher seed protein concentrations, lower seed oil concentrations, earlier maturities, more lodging, and lower yields than lines homozygous for the A81–356022 allele, which is consistent with Set 1 results. The results across the Set 2 populations indicate that the QTL-containing region can be delimited to a 3-cM interval between SSR marker Satt239 and AFLP marker ACG9b.
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
These backcross populations allowed us to more narrowly define the position of QTL for seed and agronomic traits than was possible with the experiments described by Sebolt et al. (2000). In their BC3 population, they mapped QTL for seed protein and oil concentration, maturity, plant height, seed yield, and seed weight to a 10-cM interval between the RFLP marker A515, which maps 2 cM above Satt354, and the SSR marker Satt587. They determined that the marker with the greatest effect for protein, seed yield, maturity, and plant height was Satt127, whereas the marker with the greatest effect for oil and seed size was the RFLP marker A144. On the current composite soybean map (Song et al., 2004; see also http://soybase.agron.iastate.edu/; verified 7 December 2005), A144 maps 3 cM above Satt127.
In the Set 1 populations, QTL for yield, protein and oil concentration, maturity, and seed size were mapped to an 11-cM interval between Satt614 and Satt354. We feel confident that the QTL is located in this interval since this is the only region segregating in all three populations. The Set 2 populations resulted in further refinement of the QTL position. The Set 2 data narrowed the position of the QTL for protein and oil to a region of 3 cM or less flanked by Satt239 and ACG9b.
Satt239 was generally the most significant marker across Set 1 populations for the seed and agronomic traits. However, the results from the Set 2 populations show the QTL for seed protein and oil concentrations must map below this marker, since Satt239 was not segregating in P-C609–45–2 or P-C609–45–2-2, whereas the QTL for protein and oil was still segregating in those populations. Chung et al. (2003) also mapped QTL for seed protein and oil concentrations, yield, and maturity near Satt239. Their population was developed by crossing a commercial cultivar to the high protein G. max accession PI 437088. Sebolt et al. (2000) and Brummer et al. (1997) also reported mapping protein QTL to this region in other genetic backgrounds, which suggests that the LG I QTL is common among high protein sources.
One of the goals of our study was to determine whether some or all of the negative effects associated with the G. soja allele for increased protein were due to linkage of this allele with undesirable alleles at nearby loci. If so, then such effects might be eliminated or diminished by recombination between the protein QTL and other nearby agronomic trait loci. In some populations, we did not detect significant negative effects associated with greater protein. However, the circumstances under which there was a lack of association were inconsistent. For example, yield and maturity were not significant in P-C602–15–5 but were significant in P-C602–15–6, even though these populations should have been segregating in the same region on LG I, given that they trace back to the same recombinant plant. This lack of consistency may be the result of the low statistical power of the 2004 single row tests. Further testing of these populations in replicated trials is needed to verify the single row test results.
There is evidence that points to another protein QTL in the soybean genome that is potentially homeologous to the LG I protein QTL. Shoemaker et al. (1996) studied the homeologous relationships among soybean linkage groups with RFLP markers. They found that LG I had the most markers in common with LG O, but this association was weak with only three markers shared between them. More recent work has strengthened the association between these LGs and there are now five RFLP markers in common between them (R.C. Shoemaker, personal communication, 2005). The only report we found of a protein QTL on LG O was by Specht et al. (2001). They found a relatively weak QTL on LG O in a population developed from a cross between the cultivars Minsoy and Noir 1. Although the arrangement of markers is not completely consistent between the two linkage groups, the position of the QTL on LG O in relation to markers mapping to both linkage groups is consistent with this QTL being homeologous to the LG I QTL. However, more work is needed to determine if these regions are actually homeologous.
Additional research is also needed to identify more markers in the region between ACG9b and the cluster of markers near Satt239. The identification of more markers in this region will provide further delineation of the location of the QTL segregating in these populations. These fine mapping efforts will also provide the genetic information needed in the eventual cloning of this major seed composition QTL.
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Received for publication July 5, 2005.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Burstin, P. Marget, M. Huart, A. Moessner, B. Mangin, C. Duchene, B. Desprez, N. Munier-Jolain, and G. Duc Developmental Genes Have Pleiotropic Effects on Plant Morphology and Source Capacity, Eventually Impacting on Seed Protein Content and Productivity in Pea Plant Physiology, June 1, 2007; 144(2): 768 - 781. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
