| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Agronomy, 1575 Linden Dr., Univ. of Wisconsin-Madison, Madison, WI 53706 USA
b Dep. of Plant and Soil Science, Montana State Univ., Bozeman, MT 59717 USA
c Mycogen Seeds, Box 289, Hwy. 75 North, Breckenridge, MN 56520 USA
wftracy{at}facstaff.wisc.edu
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
and ear height
but not ear weight. AFLP variation was correlated with ear weight
. While initial isolation of some of the P39 strains was due to a mutation at a single gene, the amount of molecular variation was unexpectedly high. The molecular variation and changes in combining ability indicates the occurrence of alterations throughout the genome.
Abbreviations: RFLP, restriction fragment length polymorphism • AFLP, amplified fragment length polymorphism
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Strain C30, which was isolated as a single gene mutation from P39 (Singleton, 1943), has been the most extensively studied. C30 was selected from P39 on the basis of distinct differences in plant morphology. C30 is similar in architecture to P39 except that all plant parts are smaller and C30 flowers earlier. These differences were found to be due to the reduced gene (rd1). Singleton (1943, 1944a, 1947) observed significant high parent heterosis for yield and plant height when C30 was backcrossed to P39 and Lachman (1971, 1974) observed high parent heterosis for plant height in the same cross. Singleton thought this was an example of single gene heterosis and that rd1 was responsible (1943). However, upon selfing the F1 of C30 by P39 Singleton (1944b, 1947) observed "several distinct variations," all of which segregated in "good monogenic ratios." A diminutive form of C30, Pee Wee, also was isolated (Lachman, 1971). Pee Wee has since been lost.
Sprague and Suwartardon (1975) reported a very similar case in field maize. They isolated a plant with a reduced phenotype from a stock derived from a doubled monoploid. The new phenotype was true breeding and ascribed to a recessive mutation. It was not determined if this mutation was at the rd1 locus. Upon backcrossing to the parent, they observed significant heterosis for yield plant height and other traits (Sprague and Suwartardon, 1975). Using a generation mean analysis, they determined that heterosis was due to more than one gene.
Molecular markers, including RFLPs and AFLPs, offer a virtually unlimited source of genetic markers in maize (Helentjaris, 1987; Vos et al., 1995). Transposable elements are an additional source of genomic markers in maize. Although transposable elements are noticeably active only in certain stocks, multiple sequences that hybridize to probes for Mu, Ac, and Spm are found in all maize stocks (Chandler et al., 1986; Doring and Starlinger, 1986; Talbert et al., 1989). Two cloned Mu-hybridizing sequences have been shown to be ancient and apparently stable insertions (Talbert and Chandler, 1988). Most sequences that hybridize to Ac and Spm are thought to be inactive deletion derivatives of the intact elements (Doring and Starlinger, 1986). Thus, probes for these elements may allow monitoring many genomic segments for genomic diversity with a single Southern blot hybridization.
Numerous attempts have been made to predict F1 performance on the basis of genetic distance as determined by protein- or DNA-based markers (for reviews see, Lee, 1995; Melchinger, 1999). In maize, the predictive ability of markers is strongest when studies include both intra-group and intergroup crosses (Melchinger, 1999). However, the predictive ability of the markers is much weaker when only intergroup crosses are made among unrelated materials (Melchinger, 1999).
This study was conducted to determine the amount of molecular divergence among seven strains of P39 based on polymorphisms detected with transposable element probes, RFLPs, and AFLPs. A second objective was to determine the association between molecular divergence and F1 performance in crosses among the P39 strains.
|
Materials and methods |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Laboratory Procedures
Southern blots containing EcoRI/HindIII double-digested DNA from all seven strains were probed with four clones that hybridize to multiple sequences in the maize genome following the procedure of Chandler et al. (1986). These included pDTE1, which is specific for the terminal inverted repeat of Mu elements (Chandler et al., 1986), a 1.2-kb PvuII/PstI fragment from the Ac plasmid pH3.16Ac (Fedoroff et al., 1983), an Spm specific plasmid pEco1.25 from R. Schmidt, Univ. California, San Diego (Schmidt et al., 1987), and a 1.0-kb HindIII fragment from the plasmid pMAc1 used as a probe for the multigene maize actin family (Meagher et al., 1983). Inserts were removed from the plasmids and labeled by means of a random hexamer primer reaction (Feinburg and Vogelstein, 1983).
Five strains, P39a, C30, IP39, P39M94, and P39M96, were examined for RFLPs by means of clones that hybridize to single copy nuclear DNA sequences. P39b and NewP39 were not included in the RFLP study because of resource constraints. The initial purpose of the RFLP analysis was a phylogenetic study (Gerdes and Tracy, 1994) and not all inbreds in our collection were included. The RFLP analysis was performed at Agrigenetics Company (Madison, WI) following the procedures described by Gerdes and Tracy (1994). The DNA for the analysis was obtained by bulking DNA from 15 to 20 seedlings of each inbred. The DNA was then digested with EcoR1, electrophoresed in agarose gels, and transferred to nylon filters. The clones were hybridized with the filters (Gerdes and Tracy, 1994). The resulting autoradiograms were scored to determine RFLP diversity among strains. Seventy-one covering the genome were used (Murray et al., 1988; Gerdes, 1992; Gerdes and Tracy, 1994). The clones used in this study were developed by the University of Missouri-Columbia, Brookhaven National Laboratory, and Agrigenetics Co. (Mycogen). Seven of these clones revealed two strong bands per inbred, that have been mapped to different chromosomal locations resulting in a total of 78 genomic regions assayed (Murray et al., 1988; Gerdes, 1992; Gerdes and Tracy, 1994). Markers that revealed two intense bands per strain were given a value of one if they shared at least one band and a value of zero if they did not share any common bands (Gerdes and Tracy, 1994).
The AFLP analysis of P39a, C30, IP39, P39M94, New P39 and P39M96 followed protocols supplied with Life Technologies AFLP Analysis System I Starter Primer kit (Grand Island, NY) with modifications as described by Burkhamer et al. (1998). P39b was not included because we no longer had viable seed. Equal volumes of 5 unit µL-1 Taq DNA polymerase was substituted for 1 unit µL-1 Taq DNA polymerase recommended in the protocols. A total of 25 selective amplification reactions were conducted to generate 671 polymorphic bands. The presence of a band resulted in a value of one and the absence resulted in a value of zero. The Corn Belt Dent inbred B14 was included in the AFLP analysis for comparison. Duplicate AFLP samples were not scored for these lines, however they were tested as a subset of a larger experiment involving 30 wheat lines, which included 3 blind repeats. Repeatability for these three lines was 0.92 (Talbert, Martin, and Smith, 1998, unpublished data).
Field Experiment
The seven strains were diallel crossed and seed of reciprocal crosses were bulked. Seed of the 21 resultant hybrids was planted May 1988 and May 1989 at Madison, WI, in a Plano silt loam soil (fine-silty, mixed, mesic Typic Argiudolls). In both years, the experimental design was a randomized complete block with eight replications. Plots consisted of a single row 5.3 m long with 0.76 m between rows. Plots were overplanted and thinned to 49 400 plants ha-1. Both years were dry and plots were irrigated. Plant height, from soil surface to the ligule of the leaf subtending the tassel, and ear height, from the soil surface to the ligule of the leaf subtending the ear, were measured on 10 competitive plants per row prior to harvest. A competitive plant was one that was bordered by normal vigorous plant on either side. Competitive plants were harvested after physiological maturity, approximately 50 d after flowering. Ten ears were harvested from each plot, dried at 60°C to constant moisture and weighed.
Statistical Analysis
Ear weight and plant and ear height data were partitioned into effects due to years, replications, and hybrids, by the analysis of variance. Year and replication effects were considered random while hybrids were fixed. Phenotypic correlations were used to examine relationships among agronomic traits and molecular marker classes. For the probes the number of polymorphisms between pairs was used in the correlations.
|
Results |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
|
AFLPs between inbred pairs ranged from 3 (0.5%) to 145 (22%). All strains but P39b were included, resulting in 15 pairwise comparisons. The percentage polymorphisms between B14 and each P39 strain ranged from 82 (547) to 87% (580).
When comparing P39a and P39b, no polymorphisms were detected for transposable element and actin probes. These lines were not morphologically distinct strains but had been maintained separately for over 17 generations. New P39 and IP39 were not polymorphic for the transposable marker or actin probes used in this study despite separate classification due to morphological differences. New P39 differed from IP39 at only three of over 670 (0.5%) AFLP markers. Strain C30 was the most divergent from P39a on the basis of the number of polymorphisms detected by transposable element patterns. C30 and P39a differed by 16 polymorphisms, while all the other lines differed from P39 by four to eight polymorphisms. On the basis of RFLPs at the 78 single copy loci comparisons, P39M96 was most divergent from P39a followed by C30. The comparison of C30 and P39a revealed the highest number of AFLP polymorphisms (145, 22%).
The crosses differed significantly in ear weight and plant and ear height. Ear weight ranged from 0.49 kg for New P39 x IP39, to 0.89 kg for C30 x P39. The shortest plant height (131 cm) and ear height (55 cm) were from the crosses P39a x P39b and New P39 x IP39, respectively. Plants from the cross P39M94 x P39M96 were the tallest (156 cm) and had the highest ears (72 cm).
Positive and significant correlation coefficients were observed for the level of variation among the three transposable elements (Table 2) . Among the seven inbreds, diversity as measured by Mu, Ac, Spm, actin, and total multiple sequence probes were positively correlated with 10-ear weight (Table 2). Actin sequence polymorphism was also correlated with plant height. RFLPs at the 78 single copy loci were correlated with plant height and ear height (Table 2). The correlation coefficient for the RFLPs with 10-ear weight was small. Only small and insignificant correlations were observed between the RFLPs and the transposable elements (Table 2). The number of differences between pairs calculated from RFLPs and AFLPs was not correlated. This is contrary to results in both maize and barley which showed high correlations between RFLPs and AFLPs in estimating genetic distance (Russell et al., 1997; Melchinger et al., 1998). However, those studies included a much wider range of germplasm. AFLP variation was correlated with the number of polymorphisms at all the transposable element probes, but not with the actin probes. AFLP variation was also highly correlated with ten ear weight but not with plant or ear height.
|
|
Discussion |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
The causes of molecular variation are not clear, and a number of factors may be involved. Residual heterozygosity in the original P39 strain is not a sufficient explanation. The original P39 was derived by self-pollinating a single plant, thus no more than two "alleles" are expected at a locus. RFLP data revealed seven cases of trimorphic loci. Furthermore, P39a differs from C30 and P39M96 at approximately 18% of the RFLPs an amount to high to be explained by residual heterozygosity in an inbred that had been self-pollinated six times before it was first testcrossed (Smith, 1933).
Pollen contamination or outcrossing is always a possibility in any maize breeding nursery and would result in increased phylogenetic distance determined by any type of parameter. In this instance, outcrossed plants due to pollen contamination are unlikely to have occurred for a number of reasons. First, the P39 strains are morphologically similar with the exception of the specific morphological differences for which they were identified. C30, the most extreme type, is a shorter version of P39, but similar in shape and proportion (Singleton, 1947). IP39 was identified because its kernels were slightly more narrow. Second, in the case of C30, C30 was identified because of its smaller stature. If it were an outcross, it would not have been shorter than the inbred parent P39. Upon backcrossing to P39, the reduced stature was found to be due to a single gene mutation, rd1. C30 was the original source of rd1 so it would appear unlikely that it came from an outcross. Third, five of these strains were included in a phylogenetic study of sweet corn germplasm (Gerdes and Tracy, 1994). In that study, all the P39 strains formed a discrete cluster. One non-related inbred was included in that cluster. Fourth, the other cases are not as well documented as C30 but all the P39 mutants were isolated by experienced breeders all of whom identified these as true breeding variants rather than vigorous segregating outcrosses. Any crosses to unrelated material would have been recognized as F1 hybrids, on the other hand outcrosses to P39 strains might not be recognized as outcrosses but should reduce the amount of variation among strains.
It is interesting to speculate about the role of transposable elements in generating diversity among the strains. Our data indicate a large amount of diversity is revealed with the transposable element probes. However, apparently equal amounts of diversity are revealed with the single copy probes. For the five strains for which both single copy and transposable element data are available, transposable element probes showed 73 polymorphisms between pairwise comparisons of lines with 66 hybridizing restriction fragments. RFLP probes revealed 149 polymorphisms between pairwise comparisons of lines with 163 hybridizing restriction fragments. The number of polymorphisms detected per hybridizing fragment does not differ between these two types of probes
. Thus, our data do not suggest that Mu, Ac, and Spm are actively transposing in the P39 strains in that they show no greater variability than the random single copy DNA probes. Likewise, when both RFLP and AFLP data are converted to percent polymorphisms per strain pair the range of variation is similar, 7 to 35% for RFLPs and 0.5 to 22% for AFLPs.
It is clear that P39 strains are variable at the molecular level; however, we can only speculate on the timing of the molecular changes relative to the original isolation of the strain. It is possible that the molecular changes occurred concurrently with the morphological changes. Or perhaps the majority of the molecular changes accumulated in the period between the isolation of the strain and the time of this study. In the case of the C30, it appears that the a number of changes occurred at or soon after the mutation at the rd1 locus. C30 was isolated in the mid 1930s. By 1943, enough work had been done to document heterosis in crosses between P39 and C30 and that C30 and P39 when crossed to unrelated inbreds resulted in different levels of heterosis (Singleton, 1943; Jones, 1945). But by 1944, Singleton (1944b, 1947, 1948) began to question the single gene heterosis hypothesis, and by 1947, he was quite sure other mutations had occurred. Jones (1952) also concluded that the observed heterosis of C30 x P39 was due to multiple genes.
Although the molecular variation observed does not necessarily account for the morphological differences, there does appear to be a relationship. The two P39 lines, P39a and P39b, have been maintained separately for 17 generations. However, no molecular variation was detected between these strains by means of the transposable element probes. On the other hand, C30, which was isolated as the result of a single gene mutation from P39, differs from P39a by 18 transposable element fragments, 14 RFLPs, and 145 AFLPs. It appears that morphological variation among the strains, which has been attributed to single mutations, is associated with additional genomic changes.
The fact that molecular variation exists among the strains provided an opportunity to measure the relationship between molecular diversity of parents and of F1 performance. Variation was observed for ear weight, plant height, and ear height among the crosses, the highest yielding crosses involved C30. These results confirm previous results (Singleton, 1943; Lachman, 1972, 1974). Breeding experiments indicated that the differences in yield were due to a number of genetic factors (Singleton, 1948). Molecular data support this hypothesis. Our results show a positive correlation between molecular diversity of parents as revealed by multiple sequence probes and yield of F1 progeny (Table 2). While RFLPs detected with single copy probes were strongly correlated with plant and ear height, no association was observed between these RFLPs and yield of F1 progeny. However, the reverse was true for AFLPs, which were strongly correlated with 10-ear weight but uncorrelated with plant or ear height.
Given the apparently high mutation rate and assuming that most mutations are deleterious, it is surprising that molecular diversity and improved performance of F1 progeny are correlated. However, the level of improved performance is relatively small. Even the highest performing crosses were still performing at the level of inbreds. They did not have the yield and stature of a true F1 hybrid. Sprague and Suwartardon (1975) observed a similar situation working with field maize inbreds. A line derived from a doubled monoploid apparently due to mutation, resulted in heterosis when backcrossed to the progenitor. It was subsequently determined that a number mutations had occurred.
In summary, our results indicate that widespread genomic differences are associated with morphological changes that resulted in the isolation of these strains and these differences are sufficient to result in significant changes in combining ability. Using non-molecular analysis, many workers have observed examples of apparently similar changes in maize (Jones, 1944, 1945; Singleton and Nelson, 1945; Schuler, 1954; Sprague et al., 1960; Russell et al., 1963; Burr and Burr, 1981). It would be interesting to investigate some of those reports by means of molecular methods.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Received for publication October 18, 1999.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
