| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
USDA-ARS, P.O. Box 307, Aberdeen, ID 83210
Corresponding author (pbregit{at}uidaho.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: QTLs, quantitative trait loci • DH, doubled haploid • gfw, gram fresh weight
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
One successful approach to barley transformation has been simply to utilize particular genotypes that are amenable to the requirements of transformation, such as Golden Promise (Wan and Lemaux, 1994) or Igri (Jähne et al., 1994). Another approach has been to modify the protocols used for culturing barley tissues; for instance, optimizing copper concentrations and autoclaving procedures have each been shown to increase the recovery of green plants by a factor of two (Dahleen, 1995; Bregitzer et al., 1998a). Increased copper levels combined with particular combinations of phytohormones, and the resultant improvement in regenerability, have been instrumental in the transformation of once-recalcitrant barley genotypes, and in increasing the efficiencies of transformation of amenable genotypes (Lemaux et al, 1999).
A genetic approach to improving plant regeneration would be selection for the accumulation of favorable alleles for regenerability into a single, presumably superior, genotype. Plant regeneration from cultured tissues has been shown to be under genetic control in a number of species, and genetic markers associated with plant regeneration have been identified in cereals such as barley (Komatsuda et al., 1995; Mano et al., 1996), wheat, (Triticum aestivum L.; Ben Amer et al., 1997), rice, (Oryza sativa L.; He et al., 1998), and maize (Zea mays L.; Armstrong et al., 1992). Several examples of genetic manipulations of regenerability exist. Regenerability characteristics have been transferred via hybridization and selection in maize (Armstrong et al. 1992) and in barley (Komatsuda et al., 1995). In the study by Komatsuda et al. (1995), a single locus was identified; however, a subsequent study identified three additional loci influencing plant regeneration from barley callus (Mano et al., 1996). Further work to identify and study the loci linked to these QTLs may lead to a greater understanding of the physiological processes involved in the growth and differentiation of somatic tissues. Syntenic relationships among cereal species (Van Deynze et al., 1995) suggest that the results of such studies will be broadly applicable to many species.
The objectives of this study were to identify QTLs affecting green and albino plant regeneration from barley callus. We identified at least two previously unreported QTLs for morphogenesis in cultured barley tissues, and provide confirmation of three of the four previously reported QTLs (Komatsuda et al., 1995; Mano et al., 1996). In addition, predictive models were developed and tested for use in marker-facilitated selection schemes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Callus Initiation and Plant Regeneration
Detailed descriptions can be found in Bregitzer et al. (1995). Briefly, callus was initiated by placing five immature zygotic embryos (1–3 mm in length) scutellum-down on a modified MS (Murashige and Skoog, 1962) medium supplemented with 3 mg L-1 (6.8 µM) 2,4-D. Developing calli were maintained in dim light (approximately 1 µmol m-2 s-1) provided by shaded fluorescent lamps. Germinating zygotic meristems were removed from the developing calli 1 to 3 wk after the embryos were placed on the medium (initiation). Four weeks after initiation, embryogenic-appearing sectors from the four most vigorously growing calli were subcultured onto fresh medium and subsequently maintained by biweekly subcultures. Eight and 12 wk after culture initiation, sectors from each callus were transferred to regeneration media under brighter light (approximately 5 µmol m-2 s-1) provided by unshaded fluorescent lamps. The fresh weight of callus transferred to regeneration medium was recorded.
Statistical Analyses and Model Development
The regeneration response of each DH line was determined by three separate evaluations. For each evaluation, the regeneration response was the arithmetic mean of the number of plants per g fresh weight (gfw) of callus that were recovered from each of 10 to 15 petri plates (Evaluations 1 and 2) or 15 to 20 petri plates (Evaluation 3). Data on plant regeneration were examined first for differences among lines via analyses of variance by PROC CATMOD (SAS Institute Inc., 1999) and then for associations with markers mapped by the North American Barley Genome Mapping Project (Kleinhofs et al., 1993). For this work, a 123-marker subset was used (kindly provided by P. Hayes, Oregon State Univ., USA) in two ways. First, phenotypic and genotypic data were analyzed via Mapmaker/QTL version 1.1 (Whitehead Inst., Cambridge, MA, USA) utilizing as a criterion of significance an LOD score of 2.0 or greater. Second, associations between the 123 polymorphic markers and the phenotypic data on regeneration were examined via linear regression on a per-evaluation basis. Markers were considered as candidates for model development if they had an LOD score of at least 2.0, and/or if they were shown to be significantly associated (P < 0.02) with regenerability in at least two of the three evaluations.
These data formed the dependent data set from which models describing regeneration characteristics could be developed. Potential models were developed for main effects and two-way interactions by the Mallow's Cp selection function of the PROC REG procedure within SAS (SAS Inst., Inc., 1999). Two "best" models were chosen (one for the prediction of green plant regeneration, and one for the prediction of albino plant regeneration).
Model Testing
These models were then tested for their predictive abilities using an independent data set consisting of the regeneration responses of an additional 25 DH lines, randomly chosen from the same Steptoe/Morex DH population. The Steptoe and Morex parents and the three DH lines 12, 20, and 40 were included in these analyses to facilitate comparisons of these data to that obtained from the 77 DH lines. The regeneration responses of the 25 DH lines were measured in two evaluations as described above, and compared with the predicted responses for these lines generated by the models chosen for green and albino plant regeneration.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The regeneration responses of 8- and 12-wk-old callus were highly correlated, and only the responses of eight-week-old callus (Fig. 1) were considered in the following analyses. The numbers for green plant regeneration per gfw of callus ranged from 0.1 to 44.5, with a mean of 6.7 per gfw callus. Transgressive segregation occurred also for albino plant regeneration. Albino plant regeneration ranged from 0 to 18.5 per gfw callus, with a mean of 2.0 per gfw callus. Correlation between green and albino plant regeneration was r = 0.52 (P = 0.001); such a relationship may be expected based on the assumption that both green and albino plant regeneration have common physiological origins.
|
|
|
|
The relatively low LOD scores obtained for several QTL peaks were unexpected. Previous studies in our laboratory had shown large experimental errors in the measurement of plant regeneration responses, and the decision was made to dedicate resources to multiple measurements on a subset of lines rather than single measurements of the larger 150 line population that was available. In retrospect this decision may have degraded the quality of our data, since single measurements of all 150 DH lines (Mano et al., 1996) produced generally higher LOD scores (for those QTLs detected both by this study and by Mano et al., 1996; see Comparisons to Previously Published Studies below). A theoretical treatment of this issue showed that increasing the number of individuals in the reference population, and therefore the number of potential recombinant genotypes, can be more important than increasing replication of phenotypic measurements (Knapp et al., 1990). An alternative explanation for low LOD scores may have been the observed variability between the multiple growth chambers used for plant growth, which would have increased the experimental errors of our models. Nevertheless, the data generated by our experiments appeared to be useful, and a careful reexamination of the data by regression analyses provided confirmation that the use of an LOD value of 2.0 was justified.
Model Development and Testing
Several putatively useful models were suggested on the basis of multiple regression analysis, and tested for their predictive abilities. The simplest model that appeared to be most predictive for green plant regeneration included four of the eight identified QTLs as main effects (psr129, abg019, dor4a, and abg705—see Table 2). Two additional QTLs (ksud17 and wg908) were included in two locus interactive effects, and resulted in a model that explained approximately 62% of the observed variability 
. The model that appeared most predictive of albino plant regeneration included both identified QTLs, and it predicted approximately 12% of the observed variability 
. The parameter estimates for significant main effects and interactions are shown in Table 2.
|
The range of regeneration responses for these DH lines were similar to those of the 77 DH lines used to generate the models, except that no lines were identified that had the very high levels of regenerability seen in several members of the 77-line data set. Green plant regeneration varied significantly (P < 0.05) and ranged from 0.3 to 19.7 plants per gfw callus, with a mean of 7.4 +/- 1.7. Albino plant regeneration varied significantly, and ranged from 0.1 to 2.5 plants per gfw callus, with a mean of 0.8 +/- 0.2.
For green plant regeneration, the predicted and observed means of the predicted-high group (n = 8) were 11.9 and 9.2, respectively. For the predicted-low group (n = 7), predicted and observed means were 1.2 and 4.6, respectively. For both the predicted-high and predicted-low groups, the observed means fell outside of the 5% confidence interval (CI) for the overall mean of the 21 DH lines (CI = 5.9 to 9.1).
For albino plant regeneration, the relatively simpler model generated only four predicted values. The highest (4.2, n = 5) and lowest (0.8, n = 6) predicted response groups were chosen for comparisons to observed responses. The observed means of the high and low predicted response groups were 1.2 and 0.9, respectively. The observed mean albino plant regeneration per gfw callus for all 22 lines was 0.8 +/- 0.2 (5% CI = 0.6 to 1.0). Only the predicted-high group mean observed response fell outside of the confidence interval for the overall mean.
Comparisons to Previously Published Studies
Komatsuda et al. (1991)(1993, 1995) and Mano et al. (1996) identified QTLs for green plant regeneration from barley callus. Mano et al. studied the same population (Steptoe/Morex-derived DH lines) as was studied for this report, whereas Komatsuda's group studied Japanese cultivars that are only distantly related to the U.S.-developed Steptoe and Morex cultivars. Comparisons of QTLs identified by Mano et al., Komatsuda et al., and in this study, were facilitated by reference to the consensus maps developed by Langridge et al. (1995) and by Qi et al. (1996), by the map produced by the North American Genome Mapping Project (Kleinhofs et al., 1993), and by data available on the Graingenes website at http://wheat.pw.usda.gov (verified September 7, 2000).
The QTLs for green plant regeneration linked to abg461-psr129 (7H) and abg705-abc483 (5H), and the QTLs for albino plant regeneration linked to glb1-abc494 (1H) and abc171 (3H), have not been previously identified. Although Mano et al. (1996) did not report the QTL linked to abg705-abc483, their reported LOD scores neared 2.0 in this region (see Fig. 5, Mano et al., 1996). It is interesting that the QTL linked to abg461-psr129 was not identified by Mano et al.; the relatively high (3.42) LOD score and importance of this allele in the predictive model suggests that this QTL has a major effect on green plant regeneration. Such a discrepancy between studies of the same population (Steptoe/Morex) may be attributed to differences in how plant regeneration was measured; Mano et al. (1996) measured plant regeneration as the percent of calli with plantlets, while this study measured plant regeneration as the numbers of plantlets per g fresh weight of callus. In our experience, these two measurements have not been tightly correlated (because of greater variability for total numbers of plantlets than for the percentage of calli with plantlets). Thus, we believe that measuring total numbers of plantlets provides a better characterization of regenerability.
The QTL for green plant regeneration linked to abg019-abc162 on chromosome 2H (this study) is probably analogous to the previously identified QTLs Qsr1 and Shd1. Komatsuda and colleagues (Komatsuda et al., 1991, 1993, 1995) studied shoot regeneration in the progeny of a cross between a two-rowed and a six-rowed cultivar, and described the locus Shd1 and its close linkage to the V-v locus that controls the fertility of lateral florets (2- versus 6-rowed heads); this region is within 10 cM of the QTL linked to the interval abg019-abc162 identified in this study. Qsr1 (Mano et al., 1996) was placed between the markers abg316c and abc167b; abc162 (this study) is located within this interval. The detection of this QTL in the progeny derived from six-rowed parents (Steptoe and Morex; this study, and Mano et al., 1996) suggests that the QTL for regenerability is not the V-v locus.
This study identified QTLs linked to the interval dor4a-abg471 and to ksud17; these appear analogous to the markers identified by Mano et al. (1996) as Qsr2 and Qsr3 on chromosomes 3H and 6H, respectively. The QTL associated with albino plant regeneration detected in this study (abc171) is near the dor4a-abg471 interval identified in this study for green plant regeneration (approximately 17 cM; the confidence interval for dor4a-abg471 extends to abc171), and is coincident with Qsr2. Thus, these QTLs probably are not independent of each other and represent the same locus.
Mano et al. (1996) did not detect the QTLs identified in this study in the intervals abc166-abc174 on chromosome 3H nor abg500b-abg472 on chromosome 4H. These intervals were not included as significant effects in our predictive model and may be artifacts of experimental error.
There appears to be a QTL on chromosome 5H for which the location has not been adequately determined. In this study, a QTL was detected on chromosome 5H the interval wg908–abg495a, which at first glance appears unlinked to the interval cdo57b–msrh (Qsr4, Mano et al.), but these regions may not represent different QTLs. First, this comparison is complicated by the 40 cM gap between cdo504 and abg473. Second, the confidence interval (data not shown) for wg908-abg495a extended beyond cdo504 and nearly overlapped the confidence interval reported for Qsr4; the peak for Qsr4 was broad (see Fig. 5, Mano et al., 1996). Finally, this study identified several other markers in this disputed region for which LOD scores neared significance. Additional data are needed to clarify the number and position(s) of QTLs for green plant regenerability on chromosome 5H.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Only two QTLs were identified for albino plant regeneration. One (abc171) was linked to a major QTL (Qsr2) for green plant regeneration, suggesting an association with some basic aspect of morphogenesis that is not specific to the creation of albino plants; it probably does not represent an independent QTL. The QTL linked to the interval glb1-abg494 has not been previously identified; we have termed this QTL Asr1 (albino shoot regeneration). The predictive model based on these QTLs explained only 12% of the observed variability for albino plant regeneration, suggesting that environmental factors played a relatively more major role in the regeneration of green plants than did genetic factors, and that this study was not able to accurately measure albino plant regeneration. Thus, the model describing albino plant regeneration may be of little practical utility for marker facilitated selection without further development.
These results demonstrate the feasibility of marker-facilitated selection for enhanced green plant regeneration. Cultivars such as Morex, which represents a commercially important germplasm pool that has poor regenerability characteristics, could be developed into genotypes that are elite for both regenerability and for commercial quality and agronomic characteristics. Additional efforts to more precisely locate and describe the nature of the genes these QTLs represent may lead to a greater understanding of the processes which regulate morphogenesis in barley and related members of the Gramineae family.
|
ACKNOWLEDGMENTS |
|---|
Received for publication January 25, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. P. Jauhar, S. S. Xu, and P. S. Baenziger Haploidy in Cultivated Wheats: Induction and Utility in Basic and Applied Research Crop Sci., May 11, 2009; 49(3): 737 - 755. [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 | |||
