| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Agronomy and Range Science, Univ. of California, Davis, CA 95616-8515, USA
b Inst. de Recursos Biológicos, INTA, Villa Udaondo, (1712) Castelar, Buenos Aires, Argentina
* Corresponding author (jdubcovsky{at}ucdavis.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: ANCOVA, analysis of covariance • ANOVA, analysis of variance • CS, ‘Chinese Spring’ • G x E, genotype x environment • GSP, grain softness protein • PINA, PINB, puroindoline a and b proteins • RCBD, randomized complete block design • RFLP, restriction fragment length polymorphism • SKCS, single kernel characterization system
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Different approaches have been used to characterize the natural variation observed in this character. Genetic analyses have shown that endosperm texture is primary controlled by the Hardness (Ha) locus on the short arm of chromosome 5D (Mattern et al., 1973; Law et al., 1978). It is a simply inherited character (Symes, 1965), and although the main locus is referred as hardness, softness is in fact the dominant trait.
Biochemical studies suggested that differences exist in the binding strength of starch granules with the surrounding protein matrix between soft and hard wheat varieties (Pomeranz and Williams, 1990). A starch surface-associated protein (Mr 15kd) was found at high levels in soft wheat and at relatively low levels in hard wheat (Greenwell and Schofield, 1986). This protein, referred to as friabilin, became a putative marker for softness. The accumulation of friabilin in seeds was found to be controlled by the short arm of chromosome 5D, reinforcing the idea that friabilin could be the product of the Ha locus (Jolly et al., 1993).
Friabilin was found to be a composite of related lipid-binding proteins including the basic cysteine–rich proteins puroindolines a (PINA) and b (PINB) and the grain softness protein family GSP-1 including GSP-1a, GSP-1b and GSP-1c (Gautier et al, 1994; Rahman et al., 1994). The main components have been shown to be the puroindolines a and b. Molecular studies revealed that mutations in the puroindoline gene sequences were present in all hard-textured wheats. Either a deletion resulting in the complete lack of PINA protein, or single -nucleotide mutations resulting in a modified amino acid sequence or null expression of the PINB protein were shown to be inseparably linked to hard-textured grains in surveys of American and European wheats (Giroux and Morris, 1997; 1998; Lillemo and Morris, 2000; Morris et al., 2001). These results suggest a direct role for the effect of the puroindolines on grain texture. No clear evidence on the role of Gsp-1 genes on grain texture has been obtained so far.
Genes encoding these proteins have been mapped on the distal part of chromosome 5DS completely linked to the Ha locus. Puroindoline loci, Pina-D1 and Pinb-D1, have been detected only on chromosome 5D, while homoeologous Gsp-1 loci have been conserved in all the three genomes, on chromosomes 5A, 5B and 5D (Dubcovsky and Dvorak, 1995; Jolly et al., 1996; Sourdille et al., 1996; Giroux and Morris, 1997). Puroindoline genes from chromosomes 5A and 5B are present in the diploid donors of the A and B genomes but have been deleted in polyploid wheats (Tranquilli et al., 1999; Gautier et al., 2000). The Ha-related genes may act together to affect grain softness, and it was observed that genotypes having the deletion at Pina-D1 were harder than those having the single nucleotide mutation in Pinb-D1 (Giroux et al., 2000). In consequence, it is possible to speculate that the current range of grain textures available in hexaploid wheat could be expanded to harder textures if deletions for both puroindoline genes were combined in one genotype or to softer textures if active genes from the diploid species were introduced. The main objective of our study was to test these two hypotheses. An additional objective was to test the effect of the Gsp-A1 and Gsp-B1 loci on grain texture.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental Procedures
In order to detect natural deletions at the studied loci, restriction fragment length polymorphism (RFLP) markers were used to characterize the chromosome substitution lines listed above. Nuclear DNA was isolated from leaves, following the procedure described by Dvorak et al. (1988). Southern blots and hybridizations were performed as described by Dubcovsky et al. (1994). Sample DNAs were digested with Alu I, Bam HI, Bgl II, Dde I, Dra I, Eco RI, Eco RV, Hae III, Hind III, Hinf I, Mbo I, Pvu II, Rsa I, Sty I, Taq I, and Xba I restriction enzymes and screened for polymorphism at Pina-D1, Pinb-D1, and Gsp-1 loci with clones pTa31 (Sourdille et al., 1996), pTam19B2 (Tranquilli et al., 1999), and pGsp (Rahman et al., 1994), respectively. Clone pTa31, which corresponds to the Pina-D1 cDNA, was provided by Dr. P. Joudrier (INRA, Montpellier, France). Clone pGSP was supplied by Dr. S. Rahman (Division of Plant Industry, CSIRO, Australia). Absence of the corresponding grain texture related loci on each of the deletion lines was checked by RFLP analysis.
Field Evaluations
On the basis of the differences observed by the RFLP analyses at the grain texture related loci, 10 genotypes were selected for field evaluation. They included the Timstein and Red Egyptian chromosome substitution lines, deletion lines, and the 5A/5Am recombinant substitution line. Chinese Spring was included as the control and the 11 genotypes were evaluated in two field trials performed at different environments, Tulelake and Davis, CA, USA. Tulelake is located in the North Intermountain region of California and has a Tulebasin Mucky Silty Clay Loam soil (fine, mixed, superactive, mesic Aquandic Endoaquolls), whereas Davis is in the Sacramento Valley and has a Yolo Loam soil (fine-silty, mixed, superactive, nonacid, thermic Mollic Xerofluvents).
On 20 Apr. 1999, lines were seeded in a randomized complete block design (RCBD) with thirteen blocks in the field at Tulelake. Experimental units consisted of single row plots, 3.6 m in length with rows spaced 0.3 m apart. Twenty-four seeds per plot were hand planted, which represents 
10% of the normal seeding rate for commercial plots. Preplant fertilizer was applied 1 wk before planting at the rate of 107 and 135 kg ha-1 of N and P, respectively, as ammonium phosphate (16N-20P-0K). On 11 June 1999, herbicide 2,4D-Dichlorophenoxyacetic acid, dimethylamine salt (46.8%) at the rate of 2.2 L ha-1 was used to control broadleaf weeds. On 16 June 1999, herbicide Difenzoquat methyl sulfate (31.2%) was applied at the rate of 2.7 L ha-1 to control wild oats. The experiment was irrigated as needed. Individual plots were hand-harvested on 9 September and threshed using a Vogel plot thresher.
On 11 Nov. 1999, lines were sown in Davis, in a RCBD with thirteen blocks. Experimental units consisted of two-row plots, 2.4 m in length with rows spaced 0.3 m apart. Twenty-four seeds per plot were tractor-planted. Following a crop of beans, preplant fertilizer was applied 1 wk before planting at the rate of 45 kg ha-1 of N as ammonium nitrate (33N-0P-0K). Additional fertilizer was applied 65 d after planting at the rate of 63 kg ha-1 of N as ammonium nitrate (34N-0P-K). Herbicides were not used and weeds were controlled by hand-hoeing. Individual plots were hand-harvested on 27 May 2000, and threshed using a Vogel plot thresher.
There were no fungicides applied at either location since no leaf rust infections were recorded for plots in either experiment, including CS. The dry climate at both locations and the low seeding rate did not contribute to favorable conditions for fungal diseases.
The two different experiments were harvested in different years and different locations, but for simplicity they will be referred as environments throughout the paper.
Kernel hardness index, kernel weight, and kernel moisture values were determined by a SKCS (Perten Model 4100) from the average of 300 grains. The SKCS calculates the hardness index using an algorithm developed by the USDA.
Data Analyses
Hardness data from each environment were analyzed in a RCBD with 13 blocks. Analysis of variance for hardness was performed both within environments and combined across environments. Levene's test was used to confirm homogeneity of variances between environments. In the combined model, environment, genotype x environment (G x E), and blocks within environment were considered random while genotype was considered fixed. In the combined analysis, overall differences and mean comparisons among genotypes were tested using the G x E mean square as error term. Differences between environments were tested using linear combinations of mean squares [(MS environment + MS error)/(MS environment x genotype + MS block within environment)] and effective degrees of freedom (Satterthwaite, 1946).
To study the potential effect of differences in kernel weight and kernel moisture on grain texture, the previous statistical analyses were repeated with the inclusion of these two measurements as covariates (ANCOVA). Homogeneity of regression coefficients among genotypes was tested before the ANCOVA analyses. Sixteen of the 286 data points were missing because of insufficient seed for the SKCS analyses. Therefore, least square means were used for the comparisons among genotypes. Least square means for genotypes were compared with the CS control using Dunnett-Hsu test. All statistical analyses were performed using SAS 8.0 (SAS Institute, 2001).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Each of the GSP fragments observed in CS was absent in one of the deletion lines 5AS-3, 5BS-6, and 5DS-2, confirming that the Gsp-1 loci are distal to the deletion points in these three deletion lines. Hybridization of the same DNAs with RFLP probes for the Pina-D1 and Pinb-D1 loci revealed only one fragment that was absent in deletion line 5DS-2. Therefore, it is possible to use deletion lines 5AS-3 and 5BS-6 to study the effect of the deletion of the Gsp-A1 and Gsp-B1 loci on grain texture. Deletion line 5DS-2 was used to test the effect of the simultaneous deletion of Gsp-D1, Pina-D1, and Pinb-D1 loci.
RFLP analyses with 16 restriction enzymes detected some polymorphisms between CS and the chromosome substitution lines analyzed in this study. Locus Gsp-A1 was polymorphic in substitution lines CS (Timstein 5A), CS (Cheyenne 5A), and CS (Red Egyptian 5A) relative to CS with restriction enzymes Bam HI, Bgl II, Eco RV, Hind III, and Xba I. Chinese Spring had a 4.5 kb Xba I fragment for Gsp-A1 while all these three substitution lines had a larger 5.5 kb fragment (Fig. 1) . No differences were detected among Gsp-B1 alleles with the selected restriction enzymes. The only difference detected for Gsp-D1 was the presence of a deletion in CS (Red Egyptian 5D). This substitution line had deletions at the Pina-D1 and Pinb-D1 loci (Fig. 1). Substitution line CS (Timstein 5D) showed a single deletion for Pina-D1. This difference between CS (Red Egyptian 5D) and CS (Timstein 5D) was used to study the relative effect of the simultaneous deletion of the three loci relative to the deletion of Pina-D1 on grain texture.
|
The translocation of the T. monococcum segment of chromosome 5AmS in the CS recombinant substitution line was detected with probes for the Gsp-Am1, Pina-Am1, and Pinb-Am1 genes using restriction enzyme Hind III. These three loci were reported to be present and tightly linked on the distal end of chromosome 5Am in another accession of T. monococcum (Tranquilli et al., 1999).
Four hexaploid varieties, currently grown in a large area of California Central Valley, and one advanced breeding line were also included in the screening membranes and are reported here because of the presence of a deletion in the Gsp-1 locus. These varieties were not included in the field trial that was limited to substitution and deletion lines in the isogenic background of CS. Cultivars ‘RSI5’, ‘Yecora Rojo’, ‘Kern’, and breeding line UC1041 showed a missing Gsp-1 restriction fragment with most of the restriction enzymes, which was assigned to the Gsp-A1 locus. This allelic variant is not present among the substitution lines (Fig. 1) and has not been reported previously in germplasm collections. No polymorphism was observed in Gsp-B1 and Gsp-D1 loci for this set of varieties. The deletion observed for Pina-D1 was also present in Yecora Rojo and UC1041 (Fig. 1). Lines selected for field evaluation and their genotypes are summarized below.
Grain Texture Variation among Genotypes
Hardness index values were significantly correlated (P < 0.01) with grain weight and grain moisture, but correlation values were small and explained only a small proportion of the variation in grain texture (grain weight r = 0.40; grain moisture r = 0.31). In order to eliminate a possible effect of the variation of these correlated traits on grain texture, the analysis of variance results were compared with analyses of covariance that included grain weight and grain moisture as covariates. The respective P values are included in the last two columns of Tables 1 and 2
.
|
|
The interaction between genotypes and environments was also significant (Table 1), but its contribution to the total variation was small compared with the genotype effect. The complete model explained 97.4% of the total variation in grain texture present in this experiment. Partition of the variance components showed that the largest proportion of the total variation was explained by the variation among genotypes (87.4%) and between environments (8.5%). Only a small percentage of the variation was explained by the variation among blocks (0.6%) and by the G x E interactions (0.8%).
The addition of the grain weight as covariate increased the proportion of the variation explained by the model to 98.1% and reduced the variation among environments to nonsignificant levels. Adjustment of the means by grain moisture did not increase the proportion of the variation explained by the model (97.4%). This was expected based on the low correlation between texture and grain moisture.
The relative order of the genotypes based on hardness index values was almost identical for both environments (Fig. 2) as expected based on the low percentage of the variation explained by the G x E interaction (<1%). Therefore, mean comparisons were performed simultaneously for both environments (Table 2).
|
30 units higher than CS hardness index (Table 2). Deletion line 5DS-2 did not differ significantly from substitution line CS (Red Egyptian 5D). The adjusted means for the hardness index values of the two lines carrying two deletions in the puroindoline genes were significantly higher in all the analyses (P = 0.02) than the values observed in substitution line CS (Timstein 5D) carrying a single mutation in the Pina-D1 gene (Table 1). A similar result was observed in the analysis for each of the individual environments.
The substitution line of the recombinant chromosome carrying the segment of chromosome 5Am from T. monococcum was in the other extreme of the hardness spectrum. This line was significantly softer than CS in all the statistical analyses (P < 0.01) including both combined and individual environment analyses. This line showed hardness index values 9 units smaller than the CS control (Table 2).
Substitution line CS (Timstein 5B) showed hardness index values slightly smaller than CS (average 6 units, P = 0.04). These differences were not significant in the combined analysis after correction with the grain weight in the covariance analysis (P = 0.21, Table 2). Differences between these two lines were significant (P < 0.01) in the individual environments in both the ANOVA and ANCOVA analyses due to the smaller error term of these analyses. Therefore, additional studies will be necessary to confirm this effect, and also to determine if this difference is caused by a different Gsp-B1 allele or by differences in other regions of the Timstein 5B chromosome.
Genotype by Environment Interactions Within Hard and Moderately Hard Classes
The G x E interactions explained only 1% of the variation in hardness index values in the analysis including all genotypes (Table 1). A different result was observed when two separate analyses were performed for the hard-texture genotypes [CS (Timstein 5D), CS (Red Egyptian 5D), and CS Deletion 5DS-2] and for the other seven genotypes.
Partition of the variance components in the ANOVA for the three hard-texture genotypes revealed that the genotypic differences between the lines with single and double puroindoline deletions accounted for 23.1% of the total variation whereas the differences between environments accounted for 28.8% of the variation. Also, a large component of the environmental variance (58.9%) was observed in the ANOVA for the seven moderately hard-texture genotypes. Genotypic differences among these lines accounted for only 18.3% of the total variation. The genetic variation component was significantly smaller in both separate analyses by texture class than in the analysis including both classes.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The differential effect of the Gsp-1 and puroindoline loci on grain texture parallels their similarities in protein sequence. Puroindoline a and b proteins are more similar to each other (67.5% similar) than to the GSP protein [57 and 58% similar, respectively; Gautier et al., 1994; Rahman et al., 1994; and our comparisons using ClustalW (unpublished data, 2002)].
Absence of the Pina-D1 transcript or mutations of the Pinb-D1 product are associated with hard-textured phenotypes (Giroux and Morris, 1998; Morris et al., 2001). We show in this study that the simultaneous deletion of both puroindoline loci results in average hardness indexes 5% higher than those observed in genotypes with individual mutations. This result suggests that mutation of one puroindoline locus is not enough to completely eliminate the effect of the other on grain texture.
It is interesting to point out that the hexaploid Red Egyptian 5D substitution in CS has an identical puroindoline composition as tetraploid wheats (no puroindoline genes present). Therefore, the Red Egyptian double deletion offers the possibility to engineer durum-like hardness in hexaploid wheat. The combination of the double puroindoline deletion with genes for yellow semolina color, white grain color, and low polyphenol oxidase activity may open the way for the development of hexaploid varieties with acceptable pasta quality.
The incorporation of additional copies of puroindoline genes in the substitution line with the recombined 5A/5Am chromosome produced the opposite effect to the elimination of puroindoline genes in the substitution lines described above. The presence of the Pina-Am1 and Pinb-Am1 loci from T. monococcum was associated with a decrease of 14% in the mean hardness index value compared with the control CS, and with the lowest hardness index value among all the analyzed genotypes (Table 2). This result indicates that the T. monococcum genes can be expressed in the hexaploid background and that they can act additively to the puroindoline genes present in chromosome 5D.
In this study, the effect of the T. monococcum genes was determined in the moderately hard germplasm CS. The same T. monococcum genes are currently being introgressed into soft wheat varieties to test their value to modulate soft grain textures.
The usefulness of the T. monococcum puroindoline genes in breeding programs might be limited if the introduced chromosome segment from this species has any detrimental effect on agronomic performance. Although this aspect should be further evaluated, in the present study no differences in grain weight were observed between the 5A/5Am recombinant line (29.2 g) and the CS control (29.5 g). We have initiated a project aimed to further reduce the 5AmS segment by additional cycles of recombination in the presence of the ph1b mutation.
The puroindoline genes present in the T. monococcum segment have been transferred to the 5A chromosome and therefore they would be easier to transfer to a tetraploid background than the puroindoline genes present on chromosome 5D. However, it is difficult to predict at this point the effect of the T. monococcum puroindoline genes on grain texture in pasta wheats, and therefore difficult to predict the possible applications of such genotypes.
The Red Egyptian and T. monococcum Ha loci containing lower and higher number of active puroindoline genes may become useful tools to modulate grain hardness within both the hard-textured and soft-textured hexaploid wheat varieties.
|
ACKNOWLEDGMENTS |
|---|
Received for publication January 7, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Bonafede, O. Chicaiza, G. Tranquilli, and J. Dubcovsky Registration of a Hexaploid Wheat Translocation Line Carrying a Short Segment of Chromosome 5Am including Softness Genes Pina and Pinb from Triticum monococcum Journal of Plant Registrations, May 1, 2008; 2(2): 165 - 166. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-C. Jing, D. Kornyukhin, K. Kanyuka, S. Orford, A. Zlatska, O. P. Mitrofanova, R. Koebner, and K. Hammond-Kosack Identification of variation in adaptively important traits and genome-wide analysis of trait marker associations in Triticum monococcum J. Exp. Bot., October 1, 2007; 58(13): 3749 - 3764. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Bonafede, L. Kong, G. Tranquilli, H. Ohm, and J. Dubcovsky Reduction of a Triticum monococcum Chromosome Segment Carrying the Softness Genes Pina and Pinb Translocated to Bread Wheat Crop Sci., March 1, 2007; 47(2): 821 - 828. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. N. Massa, C. F. Morris, and B. S. Gill Sequence Diversity of Puroindoline-a, Puroindoline-b, and the Grain Softness Protein Genes in Aegilops tauschii Coss Crop Sci., September 1, 2004; 44(5): 1808 - 1816. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. See, M. Giroux, and B. S. Gill Effect of Multiple Copies of Puroindoline Genes on Grain Softness Crop Sci., July 1, 2004; 44(4): 1248 - 1253. [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 | |||
