Efficient production of doubled haploid wheat plants by in vitro treatment of microspores with trifluralin or APM

Isolated microspore cultures from two doubled haploid (DH) lines of wheat, Triticum aestivum L., were used to develop an in vitro chromosome-doubling protocol. During the initial 24 h or 48 h of culture the microspores were treated with either of the two antimicrotubule herbicides trifluralin or amiprophos-methyl (APM) in concentrations ranging from 0.1 μM to 10μM. Untreated control cultures yielded 209 embryos per 100000 microspores, which is the equivalent of one spike. Among the regenerated plantlets 67% were green, and 15% of the flowering plants were spontaneously chromosome doubled. Treatments with both the herbicides had a significant effect on chromosome doubling, measured as the percentage of fertile regenerants. With the best combination of treatment duration (48 h) and herbicide concentration (10/μM) the percentage of fertile plants among regenerants could be increased up to 74% with APM and up to 65% with trifluralin. The largest numbers of DH plants per spike could be obtained with herbicide concentrations at 1–3 μM. Treatments with either herbicide at these concentrations resulted in an estimated average between the two genotypes of 27 DH plants per 100 000 microspores. These results demonstrate the high potential of APM and trifluralin as chromosome-doubling agents in isolated microspore cultures. The in vitro treatment integrated into tissue culture procedures will constitute an efficient method for chromosome doubling in future wheat breeding     

Cytological Evidence for Fertilization in Hexaploid Wheat × Sorghum Crosses

Crosses were made between the hexaploid wheat ‘Chinese Spring’ (2n = 42) and the diploid grain sorghum ‘S9B’ (2n = 20). Sixty-nine out of 100 florets fixed 48 h after pollination contained an embryo, an endosperm, or both, a remarkably high frequency in view of the taxonomic distance spanned by the cross. Percentages of single or double fertilization ranged from 50 % to 91 % for individual spikes. The hybrid origin of the embryos was confirmed by examining zygotes from spikes fixed 25 to 27 h after pollination. Seven of the 8 zygotes in which chromosomes were sufficiently contracted to be counted contained 21 large wheat chromosomes and 10 much smaller sorghum chromosomes. The eighth contained 21 chromosomes from wheat and 20 from sorghum. Sorghum chromosomes did not appear to be attached to the spindle in zygote nietaphases and showed no evidence of movement towards the spindle poles in the single zygote anaphase found. Embryos with two or more cells invariably contained one or more micronuclei and metaphases in embryos with three or more cells contained only 21 wheat chromosomes showing that sorghum chromosomes were rapidly eliminated. Endosperm, when present, was always highly abnormal.     

Agronomic Performance of Wheat Doubled-Haploid Lines Derived from Cultivars by Anther Culture

To determine the level of gametoclonal variation among doubled-haploid lines (DHLs) of Triticum aestivum L. developed using anther culture techniques and its effect on agronomic performance, 70 anther culture-derived DHLs of ‘Kitt’ were compared for 7 agronomic traits to 50 single-seed descent-derived lines (SSDLs) of ‘Kitt’ and to the cultivar ‘Kitt’. In a second experiment, 26 DHLs of ‘Chris’ were compared for 7 agronomic traits to 29 SSDLs of ‘Chris’ and to the cultivar ‘Chris’. Each experiment was grown as a randomized complete block design with three replications in three environments. For ‘Kitt’, the DHLs averaged significantly lower grain yields than the comparable SSDLs. For ‘Chris’, the DHLs averaged lower, but not significantly lower, grain yield than the SSDLs. In both ‘Kitt’ and ‘Chris’, the genetic component of variance for yield of the DHLs was significantly larger than that of the SSDLs indicating the presence of gametoclonal variation. The lower average grain yield of the DHLs was explained by a larger group of low-yielding DHLs than was present in the SSDLs. Six ‘Kitt’ DHLs and 3 ‘Chris’ DHLs were lower yielding than the lowest yielding SSDL, respectively. Elite DHLs and SSDLs were similar for mean grain yield performance. Though the DHLs and SSDLs were significantly different for some yield components, the affected yield component changed with the cultivar and there was no consistent effect. Significant genotype × environment interactions were detected for some traits which were caused by changes in the magnitude of differences, rather than reversals in ranking, indicating that low yielding DHLs could be culled on the basis of visual selection or single-environment testing. Hence, gametoclonal variation was induced by the anther culture techniques used in this study, tended to be deleterious for grain yield, and was sensitive to the growing environment. However, as the DHLs and SSDLs had similar expected population means based upon expected gains from selection, this gametoclonal variation should not be a major hindrance to wheat breeding.     

Regeneration of fertile doubled haploid plants from colchicine-supplemented media in wheat anther culture

The effect of colchicine added to induction medium for the production of fertile doubled haploid plants after in‐vitro anther culture was studied in wheat, Triticum aestivum L. For this, one winter and two spring wheat varieties were used. Anther cultures of the three genotypes were treated with 0.03% colchicine for 3 days at the beginning of microspore induction. Colchicine had no significant effect on anther response and embryoid production of the genotypes examined. However, in the winter wheat genotype ‘Mv Szigma’, colchicine caused a significant reduction in microspore‐derived structures. A significant decrease was also observed in plant regeneration ability of two genotypes (‘Vergina’ and ‘Acheloos’) after colchicine treatment. In addition, a significant reduction of the albinos produced was observed in all genotypes after olchicine treatment. In contrast, the regenerants obtained from the colchicine‐supplemented induction media produced significantly higher percentages of fertile plants in all genotypes. However, the level of fertility, was significantly different among the fertile plants obtained. This, together with the observation that in the case of the winter wheat variety the colchicine treatment resulted in 100% completely fertile plants with a high seed‐setting ability indicate that there is space for further improvement of the method when it is applied to spring cultivars. Finally, the increased number of seeds per 100 plated anthers obtained from all three genotypes after colchicine treatment, clearly demonstrates that the addition of colchicine to induction medium was superior to the conventional anther culture method and it could therefore be introduced into wheat breeding programmes.     

Genetic Variability for Haploid Production in Crosses Between Tetraploid and Hexaploid Wheats with Maize

Crosses were made between 14 wheat genotypes (11 tetraploid, 3 hexaploid) and a single Fl hybrid of maize that was used as the male parent. The experimental design consisted of randomized blocks with three replications. Plants were grown under controlled greenhouse conditions (day length 16 h and temperature 25 °C/15 °C, day/night). To enhance embryo survival, 2, 4-D treatment (10 mg/1) was applied to spikes 24 h after pollination with maize. Embryos were recovered from all tetraploid and hexaploid wheats at a rate of 2.09 to 26.76 per 100 pollinated florets. Haploid and doubled haploid plants were obtained from all hexaploid genotypes (T. aestivum) and from 5 of 11 tetraploid genotypes (T. turgidum var.). The most important point of these experiments was the ability to produce haploid plants from tetraploid wheat for two reasons: firstly, anther culture cannot be applied in tetraploid wheat (T. turgidum var.) due to the inefficiency of embryo formation and the high proportion of albino plants. Secondly, to date, crosses between tetraploid wheat and maize have resulted in embryo formation, but not in haploid plants.     

Improved Techniques for Haploid Production in Wheat using Chromosome Elimination

Wheat × maize and wheat × pearl millet crosses have proved efficient for haploid production using various genotypes of wheat; 22 and 27 % of florets produced embryos. In favourable conditions 6—9 haploid plants per spike were produced. The following simplifications or improvements in technique are recommended: 1. Only a single treatment with an aqueous solution of dicamba or 2,4-D (50–100 ppm) for embryo stimulation in vivo; 2. Application by spraying or dipping the spikes; 3. Application time two to four days after pollination; 4. Embryo rescue 15 to 18 days after pollination; 5. Crosses without emasculation are possible if pollination occurs 1–2 days before anthesis. More than 450 haploids and some doubled haploid (DH) lines (after colchicine treatment in vitro) were produced using these methods. No hybrid plants, chromosome additions or substitutions were found.     

Field Performance of Androgenetic Doubled Haploid Spring Wheat Lines in Comparison with Lines Selected by the Pedigree System

A total of 316 doubled haploid lines (DH) of spring wheat were compared with 621 lines selected in a pedigree system (PS) under field conditions in a breeding nursery. The lines originated from 21 crosses and the samples tested represent mean values for variables comprising the main breeding goals such as disease resistance, baking quality, and agronomic traits. In general, the DH lines were later in days to heading and shorter than PS lines of different homozygosity levels. Except for leaf rust (Puccinia recondite), the DHs were more resistant to the artificially infected diseases such as powdery mildew (Erysiphe graminis), stripe rust (Puccinia striiformis), and septoria nodorum blotch (Septoria nodorum). These differences, which were relatively small for practical selection purposes, were probably due to easier and more precise disease assessment of the homozygous DH lines. For quality characters, the DHs had a higher protein level, while the rest of the parameters were similar for both origins. The data analyzed suggest so far that androgentic doubled haploids in spring wheat are very similar to lines selected in a pedigree system in respect to all the agronomic characters tested. However, the DH lines were produced in a much shorter period of time. It is suggested that androgenetic doubled haploids be produced from F₁ hybrids and that the well-established bulk method should continue to allow selection bulk method should continue to allow selection for rare recombinants as soon as homozygosity is reached.