| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a The Samuel Roberts Noble Foundation, Ardmore, OK 73402
b Dep. of Crop and Soil Sciences, Univ. of Georgia, Athens, GA 30602-7272
c Dep. of Botany, Univ. of Georgia, Athens, GA 30602-7271
* Corresponding author (mksledge{at}noble.org)
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Abbreviations: Al, aluminum • QTL, quantitative trait loci • RFLP, restriction fragment length polymorphism
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Alfalfa productivity can be limited by Al toxicity. Aluminum is the most abundant metal found in the earth's crust, comprising up to 7% of its mass. At low pH, Al becomes soluble and available to plants, resulting in inhibition of root elongation and reduced plant growth. Al toxicity is a major factor limiting the productivity of crops throughout the world (Kochian, 1995). This is also true in the USA, where for the past century, Al toxicity associated with acid soils has been a major obstacle in alfalfa production (Rechigl et al., 1988). Surface application of lime is typically used to raise soil pH and reduce Al toxicity. Liming, however, is expensive and does not affect the pH of the subsurface soil, which can substantially reduce yield in alfalfa (Sumner et al., 1986). An alternative to liming is the breeding of plants with higher tolerance to Al (Foy, 1988).
Many crop species exhibit variable levels of Al tolerance (Putterill et al., 1991). Al tolerant genotypes of alfalfa have been identified among noncultivated diploid M. sativa subspecies (Bouton, 1996). These noncultivated, Al-tolerant genotypes could be used as donor parents to transfer Al tolerance into cultivated alfalfa. Transfer of genes from the diploid level to the autotetraploid level is possible either by somatically doubling the chromosomes in the diploid, or by taking advantage of 2n gametes, which occur at a low but regular frequency in both cultivated and wild M. sativa germplasm (Veronesi et al., 1986).
In this study, Al tolerance, measured by means of a callus growth bioassay, was used in conjunction with a genetic map to identify QTLs associated with Al tolerance in diploid M. sativa subsp. coerulea germplasm. These QTLs were first identified in three small F2 populations, and then confirmed in a larger backcross population. The effect of these QTLs on plant growth in acid, Al-toxic soil was also investigated.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Screening for Al Tolerance
The F2 and backcross progenies were screened for Al tolerance by means of a callus growth bioassay (Parrott and Bouton, 1990) as described by Dall'Agnol et al. (1996). Leaf tissue was the source of callus. Briefly, calli were established by growing on modified Blaydes medium for 28 d, as described by Parrott and Bouton (1990). Whole pieces of callus were then weighed and transferred to Blaydes medium, pH 4.0, both with and without the addition of 400 µM of Al. Callus from each individual genotype was grown on three plates with Blaydes plus Al, and three plates with Blaydes minus Al. Calli were transferred to fresh medium at 2-wk intervals, and final weights taken at the end of 8 wk. A ratio of growth on Al relative to growth without Al was scored as the Al tolerance present within each genotype. The experimental design was a complete randomization.
DNA Extraction, DNA Probes, and Southern Analysis
Methods for extraction of DNA, Southern blotting, and hybridization were as described by Brummer et al. (1991). The DNA probes were RFLPs from the UGA alfalfa genetic map (Brummer et al., 1993). Leaf material was collected and frozen in liquid nitrogen, then lyophilized for 24 to 48 h. Dried leaves were ground in a mortar and pestle with liquid nitrogen and a small amount of glass beads. DNA was extracted by a CTAB (hexadecyltrimethylammonium bromide) extraction method (Saghai-Maroof et al., 1984). Five to 10 µg of genomic DNA was digested with EcoRI, EcoRV, and HindIII. Digested DNA was separated by electrophoresis on 1.2% (w/v) agarose gels, and blotted onto GeneScreen Plus nylon membranes (PerkinElmer Life Sciences). For the F2 population, cloned cDNA inserts from an alfalfa seedling library were PCR amplified, and hexamer-labeled with [32P]dCTP, or with both [32P]dCTP and [32P]dATP. Hybridizations were carried out overnight at 65°C, followed by one wash with 2x SSC + 0.1% (w/v) SDS (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and two washes with 1x SSC + 0.1% SDS, 20 min each at 65°C. The membranes were then wrapped in plastic wrap and exposed to Kodak X-O Mat film; one intensifying screen was used for 7 d at -80°C. For the backcross population, probes were labeled with the DIG Luminescent Detection Kit [Roche Biomedical Supplies (cat. #1363514)]. Hybridizations were carried out in roller bottles overnight at 42°C, followed by two washes with 0.1x SSC + 0.1% SDS, antibody labeling, and luminescent detection, with reagents and instructions provided by the manufacturer. Membranes were then sealed in sheet protectors, and placed on autoradiographic film for 30 min to 1 h.
Marker Alleles
Alleles of markers UGAc044, UGAc141, UGAc471, and UGAc502 were named A through D, respectively. The subscripts "S" and "T" refer to the parent from which the allele was inherited, either the sensitive or tolerant parent. Markers with two alleles in one parent have an added subscript of "1" or "2" to distinguish the two alleles. Markers UGAc044 and UGAc141 each have an allele that is carried by both parents. That allele is named for the parent that is homozygous for that allele.
Soil Study
A study using soil was conducted as previously described (Smith, 1991; Dall'Agnol et al. 1996) with cuttings rather than seedlings. Thirty-two diploid, backcross genotypes were included in the soil assay. The soil, a Cecil sandy clay loam (clayey, kaolinitic, thermic, Typic, Kanhapludults), was collected from the Univ. of Georgia Plant Sciences Farm, near Athens, GA, and had the following characteristics: pHwater = 4.7; AlKCl = 0.29 meq/100g; Ca = 0.283 meq/100g; Mg = 0.073 meq/100g; P = 7 kg ha-1; and K = 104 kg ha-1. For limed treatments, lime and nutrients were added to the soil and the following test values were recorded: pHwater = 6.5; AlKCl = 0.0 meq/100g; Ca = 1.80 meq/100g; Mg = 0.56 meq/100g; P = 72 kg ha-1; and K = 240 kg ha-1. The experimental design was a randomized complete block with eight replications.
Cuttings were taken and rooted under mist for approximately 3 wk. The cuttings were inoculated with 10 mL of the Al-tolerant Rhizobium meliloti strain 59, at 107 CFU mL-1 (Dall'Agnol et al., 1996; Hartel and Bouton, 1991). At 12 wk, multiple cuttings of each genotype were washed free of soil, and the roots trimmed to 2.5 cm below the crown, and the stems trimmed 2.5 cm above the crown. The cuttings from each genotype were separated into eight classes on the basis of a visual assessment of the vigor of the cuttings, considering such factors as overall size, quantity of roots, stem number, and appearance of the leaves. The classes were then placed in blocks numbered from 1 to 8, ranging from the most vigorous class in block 1, down to the least vigorous class in block 8. Cuttings were placed in either limed, fertilized soil or unlimed, unfertilized soil. The soil and cuttings were placed in 0.72-L polystyrene cups, one cutting per cup, and watered by weight to 75% of field capacity every 2 to 3 d. After 6 wk, the plants were washed free of soil. The roots and shoots were separated, dried, and weighed.
Statistical Analysis
To identify probes associated with Al tolerance, one-way analysis of variance (ANOVA) was used. Data were analyzed one RFLP marker at a time, by means of the GLM Procedure of SAS (SAS Institute Inc., 1994). The marker genotype was used as the predictor variable and the Al tolerance score as the response variable. Means were obtained from averaging over the genotypic classes. A marker was considered to be associated with Al tolerance if there was a significant difference (P < 0.05) between marker genotype means for Al tolerance, as measured with the callus assay, or for differences in mean dry weights of roots or shoots for the soil assay, by an F-test from the type III mean squares obtained from the GLM Procedure of SAS. This probability level was chosen to enhance our ability to detect QTL across multiple experiments. The coefficient of determination (R2) was used as a measure of the magnitude of the marker association. Fisher's Protected LSD was used to differentiate among the genotypic classes using the statistical software package StatView (SAS Institute Inc., 1998). Normality of the F2 and backcross populations was tested with the Shapiro-Wilk test by the UNIVARIATE Procedure of SAS. Broad sense heritability was calculated as H = VG/[(VE/r) + VG], where VG = component of variance due to genotypes, VE = component of variance due to error, and r = replication (Wricke and Weber, 1986).
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Because of the high levels of heterozygosity present in subsp. coerulea, some marker loci had up to four different alleles. If only a single F1 had been chosen to derive an F2 population, marker alleles with a positive influence on Al tolerance could have been missed. Therefore, we chose a combination of three F1s to form multiple, small populations, from which we were able to detect effects from all alleles present in the two parents. Nevertheless, this complicated the molecular marker analysis and influenced the choice of QTL detection method. With three small populations, map construction and the use of interval analysis for QTL detection were not feasible. Single-marker ANOVA was therefore used to identify potential QTLs.
The Al tolerance distribution was not significantly different from normal in the three F2 populations, suggesting that Al tolerance is a quantitative trait (Fig. 1) . The Al tolerant parent had an Al tolerance score of 0.72, and the Al sensitive parent had an Al tolerance score of 0.48. The Al-3, Al-4, and Al-5 F1s had Al tolerance scores of 0.55, 0.61, and 0.65, respectively, near the expected mid-parent value of 0.60. The populations derived from the F1s had mean Al tolerance scores of 0.66, 0.57, and 0.65. The Al-3 and Al-5 populations had significantly higher Al tolerance means than the Al-4 population. Since Al-3 and Al-5 were derived from reciprocal crosses, these effects cannot be attributed to maternal or paternal effects. Rather, these differences probably reflect the different QTL alleles these three F1s inherited from the parents.
|
|
|
There were two UGAc141 marker alleles (Table 1). The Al-sensitive parent had the BSBS genotype and the Al-tolerant parent had the BSBT genotype. All three F1s had the BSBT genotype. The Al tolerance means for the BTBT F2 genotypes and for the BSBT F2 genotypes were not significantly different. The BTBT and BSBS F2 genotypes were also not different statistically with respect to Al tolerance. As for UGAc044, this is not consistent with either dominance or additive gene action, suggesting that this marker is also a false positive.
There were three alleles for marker UGAc471, two inherited from the Al-sensitive parent (alleles CS1 and CS2) and one inherited from the Al-tolerant parent (CT) (Table 1). The Al-sensitive parent had the CS1CS2 genotype, the Al-tolerant parent had the CTCT genotype, and the two F1s Al-3 and Al-5 had the CS1CT genotype, while the F1 Al-4 had the CS2CT genotype. The Al tolerance means for F2 genotypes CS1CS1, CTCT, CS1CT, and CS2CT were not significantly different, and were higher than the means for genotypes CS2CS2. The CS1 and CT alleles appear to have positive effects on Al tolerance.
Marker UGAc502 had four different alleles, the DS1 and DS2 alleles from the Al-sensitive parent, and DT1 and DT2 alleles from the Al-tolerant parent (Fig 3) . The sensitive parent had the DS1DS2 genotype, and the tolerant parent had the DT1DT2 genotype. The Al-3 F1 had the DS1DT1 genotype, the Al-4 F1 had the DS2DT1 genotype, and the Al-5 F1 had the DS2DT2 genotype. The DS1DS1 F2 genotypes had higher Al tolerance means than the DS1DT1 F2 genotypes, suggesting additive rather than dominance gene action. There was no significant difference in Al tolerance means between the DT2DT2 and DS2DT2 F2 genotypes, suggesting that these alleles have similar, positive effects on Al tolerance. The DS1, DS2, and DT2 alleles all appear to be positively associated with Al tolerance.
|
Al tolerance was normally distributed in the backcross population (Fig. 4) . Heritability for Al tolerance as measured with the callus bioassay was 93%, a relatively high value. This high heritability could be a result of using a tissue culture assay, with well defined growth conditions, rather than a field experiment, where environmental conditions could have resulted in large G x E interactions and lower heritability.
|
|
The marker UGAc471 genotypes were not associated significantly with Al tolerance in the backcross population (Table 2). The CS1CT genotypes, however, had significantly higher Al tolerance means than the CS2CS2 genotypes. When scored as single alleles, absence of the CS2 allele was associated with higher Al tolerance means, whereas presence of the CT allele was associated with higher Al tolerance means. It is possible that genotypes with the CTCT genotypes, which did not occur in this backcross, would have a stronger association with Al tolerance.
UGAc502 has four marker alleles. The DT2 allele, identified as a positive allele in the F2 callus study, did not occur in the backcross population (Table 2). Confirmation of this allele would require the construction of a different population. The DS1DT1 genotypes had significantly lower Al tolerance means than the other genotypes. When scored as alleles, presence of the DT1 allele is associated with Al tolerance. This allele however, also occurs in the DS2DT1 genotype, which is not significantly different than the DS2DS2 and DS1DS2 genotypes. The DS2 allele is the only allele clearly associated with Al tolerance, as measured with the callus bioassay, in the backcross population.
Soil Study
Liming of soil is the method used in the field to raise soil pH and eliminate Al toxicity. We used the backcross genotypes to determine if RFLP markers identified as being associated with Al tolerance by the callus bioassay would also be associated with traits typically used to measure Al tolerance in whole plants, such as root growth in unlimed soil.
Genotypes for the soil study were chosen on the basis of the marker alleles of markers UGAc471 and UGAc502. There were no genotypes that had only favorable or only unfavorable alleles. An attempt was made, however, to select plants that had mostly favorable or mostly unfavorable markers alleles for both markers. Al tolerance scores from the callus assay were not used to select genotypes. Therefore, marker-assisted selection, rather than selection based on phenotype, was used to select plants for this study. The purpose was not only to confirm positive alleles, but also to test the effectiveness of marker-assisted selection.
There were no alleles significantly associated with Al tolerance in the soil-based study for markers UGAc044 or UGAc141 (data not shown). This supports evidence from the F2 and backcross studies that suggest that these markers are false positives.
Marker allele CT of UGAc471 was associated positively with the dry weight of roots and tops grown both in unlimed and limed soil (Table 3). This result is consistent with both the F2 and backcross callus studies. The CS1 marker allele, which was positively associated with Al tolerance in the F2 callus study, but not in the backcross callus study, was not associated with growth in either limed or unlimed soil. The CT allele appears to be the only positive allele for UGAc471 and is the allele that should be used for marker-assisted introgression of Al tolerance into cultivars of alfalfa.
|
A ratio of growth in unlimed soil to growth in limed soil is not typically an effective measure of Al tolerance (Dall'Agnol et al., 1996; Smith, 1991). It is often the case that genotypes with poor vigor (i.e., dry matter yield) and poor growth in both unlimed and limed soil have high ratios. These ratios, therefore, may reflect low vigor rather than high Al tolerance (Scott and Fisher, 1989). Use of ratios was also not an effective measure of aluminum tolerance in our study. None of the markers showed any association with Al tolerance based on a ratio of dry weight in unlimed soil to dry weight in limed soil (data not shown).
Broad sense heritability in the soil-based assay was previously reported as 0.66 and 0.77 for root growth in unlimed and limed soil, respectively (Smith, 1991). Although Smith (1991) used M. sativa subsp. sativa and we used M. sativa subsp. coerulea, this suggests that the soil-based assay is as effective a selection strategy for Al/acid soil tolerance as the callus bioassay. Dall'Agnol et al. (1996) compared the soil and cell culture methods of selecting genotypes for Al tolerance and found a greenhouse soil-based assay to be the most efficient way to screen and select alfalfa for tolerance to acid soil and Al toxicity. The advantage of the soil-based assay is that it is less expensive, as well as less labor intensive than the callus bioassay. On the basis of the molecular marker associations observed in this soil-based study, the soil-based assay yields molecular marker associations that are consistent with those associations identified in the callus-based assays.
Efficiency of QTL experiments is difficult to measure. One method used is to sum cumulative phenotypic variance attributable to a combination of all significant QTLs (Tanksley, 1993). Where complete maps have been used for quantitative studies, these values have ranged from 10 to 95%, but typically average from 30 to 40% (Tanksley, 1993; Kearsey and Farquhar, 1998). The variation explained by markers UGAc471 and UGAc502 cumulatively accounted for 31.53% of the variation (Table 1). Single marker detection of QTLs using ANOVA, however, does not give an accurate estimation of the amount of variation explained by each QTL, since the exact location of the QTL is not known; therefore, it is not possible to distinguish between tight linkage to a QTL of small effect, and loose linkage to a QTL of large effect (Lander and Botstein, 1989). High resolution mapping of the regions surrounding the identified QTL, and interval analysis, could give a more accurate location of the QTLs and estimation of the magnitude of their effects.
Uncertainty as to the exact location of a QTL has consequences for marker-assisted breeding. The further a marker is from a QTL, the more likely it is that a recombination event could separate the QTL from the marker allele. Flanking markers tightly linked to the QTL would allow recombination between the QTL and the markers to be detected, and could also be used for marker-assisted backcrossing. It is possible that UGAc191 and UGAc471 are flanking a single QTL. The other QTL identified, however, is linked to only a single marker, UGAc502. High resolution mapping of the regions surrounding the identified QTL could be used to identify flanking markers for this QTL.
One of the two Al tolerance QTLs identified in these experiments, linked to UGAc502, was derived from the Al sensitive parent. It has been observed that unadapted germplasm, with poor trait phenotypes, often contain alleles that are actually favorable for the trait (Moncada et al., 2001). This is the basis for advanced backcross QTL breeding (Tanksley and Nelson, 1996) in which these alleles are introgressed into elite cultivars that allow for favorable expression of these alleles. While advanced backcross QTL breeding is not amenable to heterozygous, outcrossing crops such as alfalfa, this study demonstrates that these unexpressed, favorable alleles do exist in unadapted M. sativa germplasm.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. Narasimhamoorthy, E. B. Blancaflor, J. H. Bouton, M. E. Payton, and M. K. Sledge A Comparison of Hydroponics, Soil, and Root Staining Methods for Evaluation of Aluminum Tolerance in Medicago truncatula (Barrel Medic) Germplasm Crop Sci., February 6, 2007; 47(1): 321 - 328. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Robins, D. Luth, T. A. Campbell, G. R. Bauchan, C. He, D. R. Viands, J. L. Hansen, and E. C. Brummer Genetic Mapping of Biomass Production in Tetraploid Alfalfa Crop Sci., January 22, 2007; 47(1): 1 - 10. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. K. Sledge, P. Pechter, and M. E. Payton Aluminum Tolerance in Medicago truncatula Germplasm Crop Sci., August 26, 2005; 45(5): 2001 - 2004. [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 | |||
DNA; B is the Al sensitive parent M. sativa subsp. coerulea 440501-2, showing the DS1 and DS2 alleles; C is the Al tolerant parent M. sativa subsp. coerulea 464724-25, showing the DT1 and DT2 alleles; D is F1 Al-3; E is F1 Al-4; F is F1 Al-5. Population One is an F2 population derived from selfing Al-3, Population Two is from Al-4, and Population Three is from Al-5.
