| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
USDA-ARS, Forage and Range Research Laboratory, Utah State Univ., Logan, UT 84322-6300. USA. Cooperative investigations of the USDA-ARS and the Utah Agric. Exp. St., Logan, UT 84322.
* Corresponding author (kevin{at}cc.usu.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Abbreviations: AMOVA, analysis of molecular variance • AFLP, amplified fragment length polymorphism • UPGMA, unweighted pair group method with arithmetic averages
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
This widely adapted cool-season perennial grass is a complex of diploid (2n = 14), tetraploid (2n = 28), and hexaploid (2n = 42) species. On the basis of chromosome-pairing relationships in hybrids among species in the complex, Dewey (1969, 1974) concluded that the same basic genome ‘C’ (changed to ‘P’, Dewey, 1984), modified by structural rearrangements, occurred at the three ploidy levels and that the crested wheatgrasses should be treated as a single gene pool. The tetraploid complex [A. cristatum, A. desertorum, and A. fragile (Roth) Candargy] is the most widely distributed, ranging from Central Europe and the Middle East across Central Asia to Siberia, China, and Mongolia. The diploids [A. cristatum and A. mongolicum Keng] are distributed over the same general range as the tetraploids, but their occurrence is much more sporadic. Hexaploid forms are rarely reported and are found primarily in Turkey and Iran (Dewey, 1983).
Based on morphometrics and chromosome pairing, Asay et al. (1992) concluded that A. mongolicum and A. cristatum were the diploid progenitors of the standard-type crested wheatgrass A. desertorum. They further proposed that A. fragile is an autotetraploid derivative of A. mongolicum and that forms of A. desertorum could be generated from hybrids between A. fragile and tetraploid A. cristatum.
Cultivars of crested wheatgrass within the different ploidy levels have been released in North America (Asay and Jensen, 1996). However, until recently, genetic improvement in these cultivars has been restricted to selection and hybridization within ploidy levels. The most progress from interploidy breeding to date has been achieved at the 4x level. The cultivar Hycrest, which was developed from a hybrid between induced tetraploid A. cristatum and natural tetraploid A. desertorum, was released in 1984 (Asay et al., 1985). Hybridization schemes involving 6x-2x, 6x-4x, and 4x-2x have shown potential for expanding the genetic resources of 4x breeding populations (Asay and Dewey, 1979; Dewey, 1969, 1971, 1974; Dewey and Pendse, 1968; Knowles, 1955). Dewey and Pendse's (1968) data suggested that selection for improved fertility would be effective in these interploidy hybrids.
The hexaploid accession (6x-BL), PI 406442, from the former Soviet Union, is characterized by exceptionally broad leaves that maintain their green color 2 to 3 wk longer in the growing season than typical crested wheatgrass. A breeding program is in progress to transfer the ‘broadleaf’ and color retention characters into the genetic background of the tetraploid cultivar Hycrest. The objectives of these studies were to determine the range in chromosome number, meiotic regularity, molecular association, fertility, and leaf width of Hycrest/6x-BL, Hycrest*2/6x-BL, and Hycrest/6x-BL//Hycrest/6x-BL hybrids. The data will provide a basis to evaluate the potential of combining the genetic resources of diploid, tetraploid, and hexaploid crested wheatgrass through interploidy hybridization.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Cytological Analysis
Root tips were treated in an aqueous solution containing 0.05% colchicine plus 0.025% 8-hydroxyquinoline, and 25 drops 100 mL–1 of dimethylsulfoxide (DMSO), for 2 to 3 h at room temperature in darkness. They were then fixed and stained in 2% aceto-orcein at 4°C for a minimum of 3 d. The meristematic portion of the root tip was squashed in 45% acetic acid.
Pollen mother cells from the hybrid populations were preserved in Carnoy's fixative (6 parts absolute alcohol:3 parts chloroform:1 part glacial acetic acid) for 24 to 48 h, transferred to 70% ethanol, and stored in a refrigerator until analyzed. Squashed preparations of the pollen mother cells were stained with 2% acetocarmine, and chromosome pairing was analyzed at metaphase I.
Pollen Stainability and Seed Set
Spikes for pollen stainability were collected at anthesis for Hycrest/6x-BL, Hycrest*2/6x-BL, and Hycrest/6x-BL //Hycrest/6x-BL. The pollen grains were immersed in an I2KI (iodine-potassium iodide) solution, which stains starch found in viable pollen grains black or dark gray. Aborted pollen grains are shrunken and light amber colored in I2KI. A minimum of 1000 pollen grains were scored as viable or inviable for each parent and hybrid progeny. Seed set under open-pollination for the above hybrid populations was determined on 15 to 45 plants for each parent and hybrid progeny which were harvested 1 mo after anthesis. The spikes were hand threshed and seed counted to estimate plant fertility expressed as seeds per spike.
Morphological Traits
Morphological variation in the parents and the hybrid populations were measured on plant flag leaf width (mm) and flag leaf length (cm) using 15 to 45 different plants. From each plant, morphological data were collected as the mean of five measurements. All data were subjected to analysis of variance using GLM procedures as a fixed model. Mean separations were made on the basis of least significant differences (LSD) at the 0.05 probability level (SAS Institute Inc., 1999).
DNA Analysis
Genomic DNA was extracted from 11 to 14 seedlings from each of five tetraploid crested wheatgrass cultivars, two experimental tetraploid crested wheatgrass breeding populations, and other tetraploid crested wheatgrass seed lots not considered here (total of 192 plants) using the DNeasy 96 plant DNA extraction kit and MM300 mixer mill (Qiagen, Valencia, CA). All 192 DNA samples were screened with six amplified fragment length polymorphism (AFLP) primer pairs; E.ACG//M.CTA, E.ACT//M.CAA, E.AGA//M.CAG, E.AGG//M.CAT, E.AGG//M.CTA, and E.AGT//M.CTT. The AFLP technique was performed as described by Vos et al. (1995), except that the selective EcoRI primers were fluorescent labeled with 6-FAM and fractionated by capillary electrophoresis using an ABI3100 instrument (PE Applied Biosystems, Foster City, CA) with GS-400 size standards (PE Applied Biosystems) (DeHaan et al., 2002). The relative migration (molecular size) of amplified DNA fragments was determined using GeneScan (PE Applied Biosystems) and classified into allelic categories using Genographer (Benham et al., 1999). For this study, we considered 14 genotypes from each of the cultivars ‘CD-II,’ Hycrest, ‘Nordan,’ ‘Kirk,’ and ‘RoadCrest,’ and breeding populations designated I-28 and Hycrest*2/6x-BL. Genetic variation within and between populations was investigated using analysis of molecular variance (AMOVA) based on the average number of DNA polymorphisms within and between populations (Excoffier et al., 1992) and related methods of hypothesis testing based on similarity coefficients corrected for covariance (Leonard et al., 1999). AMOVA was performed using the ARLEQUIN software (Excoffier, 1992), whereas other methods of hypothesis testing based on similarity coefficients were performed using SAS programs described by Leonard et al. (1999). Pairwise comparisons of the number of DNA polymorphisms between individual plants, used for AMOVA, were calculated using SAS. Genetic relationships between individual plants were also investigated by unweighted pair group method with arithmetic averages (UPGMA) cluster analysis, using NTSYSpc software (Rohlf, 1998), based on DICE similarity coefficients equivalent to those used for comparing diversity within and differentiation among populations as described by Leonard et al. (1999). Confidence levels were obtained from 10 000 bootstrap UPGMA searches, based on MEAN distances between genotypes, using PAUP* software version 4.0b10 (Swofford, 1998).
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Hycrest/6x-BL Hybrids
Chromosome numbers ranged from 2n = 27 to 41 with 30% of the plants having a chromosome number of either 2n = 37 or 38 at metaphase I. Presence of aneuploidy in crested wheatgrass has been previously reported by Asay et al. (1986), Imanywoha et al. (1994), and Asay and Jensen (1996). Dewey and Pendse (1968) reported extra chromosomes, which they designated as B-chromosomes, in A. desertorum. Unpaired chromosomes were observed in 97% of the metaphase I cells and multivalents consisting of either 3, 4, 5, 6, or 7 chromosomes were observed in all cells (Table 1). Univalent and bivalent chromosome associations were not significantly affected by aneuploid number (data not shown) in this hybrid. There was, however, a general trend toward an increased number of univalents per cell as aneuploid numbers increased. Hybrids with 39 chromosomes had 1.2 fewer bivalents per cell than did 2n = 35 hybrids. Multivalents comprised of 6 and 7 chromosomes per cell increased significantly (P < 0.05) in the 39 chromosome aneuploids over the pentaploid hybrid, suggesting that the chromosomes are remaining as unpaired or being included in higher order chromosome associations. Mean chromosome associations for this hybrid were 2.8 univalents + 7.6 bivalents + 2.8 trivalents +0.9 quadrivalents + 1.09 pentavalents + 0.09 hexavalents +0.09 septavalents and higher. Lagging chromosomes at anaphase I occurred in over 90% of the cells scored and micronuclei were seen in 90% of the observed quartets in Hycrest/6x-BL hybrids, supporting previous reports of possible structural heterozygosity between the different ploidy levels in crested wheatgrass (Dewey, 1984).
|
|
|
Plants with 2n = 28 were meiotically more stable than those that had varying degrees of aneuploidy. In the euploid plants, univalents were observed in 14% of the metaphase I cells compared to 84% in the aneuploids. Chromosome number did not affect the frequency of bivalent formation in this hybrid combination (Table 1). However, not only was there an increase in the number of unpaired chromosomes in the aneuploids, but 67% of the cells had at least one trivalent (Table 1) compared to 7% of the euploid plants with at least one trivalent. As expected in a true autotetraploid of crested wheatgrass, quadrivalent formation was higher in the euploid plants compared to the aneuploid plants (Table 1). Multivalents comprised of more than four chromosomes were more frequent in the aneuploids than in the euploid plants, again supporting the premise that the extra chromosomes were being included in higher order chromosome associations. Mean chromosome associations in the 2n = 28 plants were 0.4 univalents + 8.5 bivalents + 0.2 trivalents + 2.6 quadrivalents + 0.07 pentavalents + 0.01 hexavalents + 0.01 septavalents and higher compared to 2.1 univalents + 7.8 bivalents + 1.2 trivalents + 1.9 quadrivalents + 0.7 pentavalents + 0.07 hexavalents + 0.05 septavalents in the aneuploids. Of the three hybrid populations, the backcross hybrid had a smaller number of lagging chromosomes at anaphase and micronuclei per quartet than did Hycrest/6x-BL and Hycrest/6x-BL//Hycrest/6x-BL, averaging 0.11 lagging chromosomes cell–1 and 0.07 micronuclei quartet–1.
The backcross hybrid (Hycrest*2/6x-BL) produced more open pollinated seed than did Hycrest/6x-BL//Hycrest/6x-BL (Table 2). Stainable pollen ranged from 40 to 91% in the backcross hybrid. Surprisingly, the level of aneuplody had no effect on percent stainable pollen (Table 2); however, there was a significant negative association between aneuploid level and reduced seed set (r = –0.47, P = 0.04). Seed set spike–1 ranged from 4.3 seeds spike–1 to 110 seeds spike–1 (Table 2). As expected, there was a significant (P < 0.01) correlation between percent stainable pollen and seed set.
The broadleaf character was readily detected in backcross populations (Table 3); however, the overall population mean declined by 3.4 mm in the backcross, but remained significantly wider (P < 0.05) than Hycrest by 3.1 mm.
Hycrest/6x-BL//Hycrest/6x-BL Hybrids
Progeny from the Hycrest/6x-BL cross, with increased leafiness and leaf width, were crossed among themselves without regard for chromosome number. In progeny from these crosses, chromosome numbers ranged from 2n = 33 to 45, with the most frequent association being 2n = 41 in 21% of the metaphase I cells. Meiotic chromosome associations were similar to those observed in the aneuploid Hycrest*2/6x-BL hybrids. At least one univalent was observed in 92% of the metaphase I cells. The most frequently observed chromosome associations were 2.5 univalents + 9.4 bivalents + 1.4 trivalents + 2.2 quadrivalents + 0.5 pentavalents + 0.3 hexavalents + 0.1 septavalents and higher order chromosome associations. Unparied chromosomes and unequal disjunction at anaphase I and the quartet stage contributed to 0.19 lagging chromosomes per cell and 0.09 micronuclei per quartet.
Pollen stainability in Hycrest/6x-BL//Hycrest/6x-BL hybrids ranged from 22 to 90% (43 plants) with an average of 62%. There was no significant difference in stainable pollen between Hycrest/6x-BL, Hycrest*2/6x-BL, and Hycrest/6x-BL//Hycrest x 6x-BL (Table 2). Seed set under open-pollination in the Hycrest/6x-BL//Hycrest/6x-BL hybrids ranged from 0 to 88 seeds spike–1 with an average of 21 ± 21 seeds spike–1. A significant correlation (r = 0.37, P = 0.01) between pollen stainability and seed set suggests that selection for increased pollen stainability will result in increased seed set. There were no significant relationships between aneuploidy and percent stainable pollen or seed set spike–1 (r = 0.21, P = 0.17 and r = –0.24, P = 0.11). However, seed set in the 42 chromosome plants was highly variable, and after three cycles of selection could not be stabilized (data not shown), which led to the discontinuation of this population at the hexaploid level.
Molecular Characterization of the 2n = 28 Chromosome Breeding Population
Due to aberrant, weak, or off-type AFLP profiles, three Nordan (Elliot and Bolton, 1970), one I-28, and one Hycrest DNA samples were removed from the analysis. Our research suggests that most of the apparent differences between Hycrest*2/6x-BL tetraploids and Hycrest were largely due to changes in allele frequency rather than the infusion of new alleles. A total of 972 fragment categories (707 polymorphic categories), containing 8471 fragments, were detected among the 14 Hycrest*2/6x-BL plants, whereas 964 fragment categories (684 polymorphic categories), containing 8018 fragments, were detected among the 13 Hycrest plants. A combined total of 1032 fragment categories were detected in both populations, including 813 polymorphic categories.
The average number of bands per plant was not significantly different between Hycrest and the Hycrest*2/6x-BL tetraploids (Table 4). However, an increased number of polymorphic fragments in Hycrest*2/6x-BL tetraploids, relative to Hycrest, is likely a balance between fixation of alleles that were polymorphic in Hycrest and the infusion of new alleles (fragments or null alleles) from the donor parent. A total of 68 fragment categories were present in 14 Hycrest*2/6x-BL plants, but absent in 13 Hycrest plants. These 68 fragment categories may reflect the infusion of new alleles in the tetraploid Hycrest*2/6x-BL population relative to Hycrest. Conversely, 60 fragment categories were present among 13 Hycrest plants, but absent in 14 Hycrest*2/6x-BL plants. These 60 fragment categories may reflect the fixation of alleles in the tetraploid Hycrest*2/6x-BL population that were polymorphic in Hycrest. Conversely, the conversion of monomorphic fragments in Hycrest to polymorphic fragments in Hycrest*2/6x-BL tetraploids may also reflect an infusion of new alleles.
|
Thus, a total of 129 potentially new AFLP alleles were gained and 106 AFLP alleles fixed in Hycrest*2/6x-BL tetraploids, relative to the 964 fragment categories (loci) detected in the original Hycrest variety. Although a substantial number of potentially new alleles (present and absent alleles) were detected in Hycrest*2/6x-BL tetraploids, we conclude that much of the differentiation between these populations (Fig. 1 and Table 5) can be attributed to allele fixation (loss) and other changes in allele frequency.
|
|
On average, approximately 17% of the DNA polymorphism (AFLP) was apportioned among the seven crested wheatgrass populations, with the remaining 83% of the DNA polymorphism maintained within cultivars. Mellish et al. (2002) reported 12% and 88% of the AFLP variation partitioned among two diploid and four tetraploid crested wheatgrass populations (including Hycrest, Kirk [Knowles, 1990], and Nordan).
CD-II and Nordan displayed the greatest and smallest average similarity coefficients within cultivars, respectively (Table 4). Conversely, CD-II and Nordan displayed the smallest and largest average number of differences within cultivars, respectively (Table 5). Interestingly, Mellish et al. (2002) ranked Hycrest, Kirk, then Nordan in descending order of genetic diversity, which is the opposite of these findings (Table 4 and 5). Although Hycrest was derived from an interspecific cross of A. cristatum/A. desertorum, this variety can be traced back to only 4 or 5 hybrid plants with additional selection pressure. CD-II was subsequently derived from only 10 Hycrest clones (Asay et al., 1997), which could certainly explain why this variety was least diverse (Table 4). In any case, the average similarity coefficient among Hycrest*2/6x-BL tetraploid hybrids was smaller, but not significantly different from Hycrest (Table 4). Conversely, the average number of DNA polymorphisms among Hycrest*2/6x-BL tetraploids appears greater than Hycrest (Table 5), but this difference was not statistically evaluated.
All pairwise comparisons of the average similarity coefficients between cultivars were significantly less than average similarity coefficients within cultivars based on the permutation test described by Leonard et al. (1999) (Table 5). However, all pairwise comparisons of the average number of DNA polymorphisms within cultivars were significantly less than the average number of DNA polymorphisms between cultivars, based on the permutation test described by Excoffier et al. (1992) (Table 5). Thus, significant differentiation among all cultivars was detectable using both measures of DNA variation (i.e., absolute number of DNA polymorphisms and similarity coefficients). RoadCrest was the most distinct cultivar in terms of average number of differences and average similarity coefficients between cultivars (Table 5), which was also apparent in the UPGMA dendogram (Fig. 1). RoadCrest was derived from two collections from Turkey (Asay et al., 1999), and has no pedigree relationship to the other crested wheatgrass cultivars in this study. Interestingly, Mellish et al. (2002) reported 11% differentiation between Kirk and Hycrest, 6% between Kirk and Nordan, and 12% between Nordan and Hycrest. Similarly, we detected 13% differentiation between Kirk and Hycrest, 9.0% between Kirk and Nordan, and 11% between Nordan and Hycrest (Table 5). In one analysis of 12 populations, Mellish et al. (2002) found that Kirk and Hycrest group together relative to Nordan. In another analysis of six populations, Mellish et al. (2002) found that Kirk and Nordan group together relative to Hycrest, which is consistent with our results (Fig. 1). The relationship of Hycrest, CD-II, and I-28 is well established based on pedigree and AFLP genotyping (Fig. 1 and Table 5). Mellish et al.(2002) also included the hexaploid crested wheatgrass cultivar Douglas in a comparison among 11 other diploid or tetraploid accessions (including Hycrest, CD-II, Nordan, and Kirk). Douglas was derived from a cross of PI 406442 and four other hexaploid accessions (Asay et al., 1995). The PI 406442 accession was used as the female parent in all crosses (Asay et al., 1995), thus Douglas should retain considerable genetic identity to the PI 406442 donor parent of Hycrest*2/6x-BL tetraploids. Among the 12 cultivars compared by Mellish et al. (2002), Nordan was most similar to Douglas. Despite the pedigree relationship between Hycrest and Hycrest*2/6x-BL tetraploids, our results indicate that the latter population is more similar to Kirk and Nordan (Fig. 1 and Table 5).
Received for publication February 16, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
of the average similarity coefficient within cultivars.
.