Plant Cell Reports

, Volume 25, Issue 4, pp 257–264

Production of doubled haploids in durum wheat (Triticum turgidum L.) through isolated microspore culture

Authors

    • Departamento de Genética y Producción VegetalEstación Experimental de Aula Dei, Consejo Superior de Investigaciones Científicas
  • M. Soriano
    • Departamento de Genética y Producción VegetalEstación Experimental de Aula Dei, Consejo Superior de Investigaciones Científicas
  • A. M. Castillo
    • Departamento de Genética y Producción VegetalEstación Experimental de Aula Dei, Consejo Superior de Investigaciones Científicas
  • M. P. Vallés
    • Departamento de Genética y Producción VegetalEstación Experimental de Aula Dei, Consejo Superior de Investigaciones Científicas
  • J. M. Sanz
    • Departamento de Genética y Producción VegetalEstación Experimental de Aula Dei, Consejo Superior de Investigaciones Científicas
  • B. Echávarri
    • Departamento de Genética y Producción VegetalEstación Experimental de Aula Dei, Consejo Superior de Investigaciones Científicas
Cell Biology and Morphogenesis

DOI: 10.1007/s00299-005-0047-8

Cite this article as:
Cistué, L., Soriano, M., Castillo, A.M. et al. Plant Cell Rep (2006) 25: 257. doi:10.1007/s00299-005-0047-8

Abstract

The objective of this work was to produce doubled haploid plants from durum wheat through the induction of androgenesis. A microspore culture technique was developed and used to produce fertile doubled haploid plants of agronomic interest. Five cultivars, one selected line, plus a collection of 20 F1 crosses between different genotypes of high breeding value were used. Studies on several factors such as pre-treatments and media components were carried out in order to develop a protocol to regenerate green haploid plantlets. Anthers were pre-treated in 0.7 M mannitol. Microspores, from anther maceration, were plated on a C17 induction culture medium with ovary co-culture. The optimum regeneration medium J25–8 was used. From 35 microspore isolations, 407 green plantlets were obtained. With this technique mature embryos were obtained. Green plants were regenerated from all genotypes used and approximately 67% of them were spontaneously doubled haploids. Some haploids and a very few polyploids plants were obtained. From the 407 plants, 275 were completely fertile and gave enough seeds to be assayed in the field. This protocol could be used complementary to or instead of the intergeneric crossing with maize as an economically feasible method to obtain doubled haploids from most durum wheat genotypes.

Keywords

Doubled haploidMicrospore cultureDurum wheatTriticum turgidum

Abbreviations

DH

Doubled haploid

2,4-D

2,4-dic-hlorophenoxyacetic acid

BAP

6-Benzylaminopurine

KIN

6-Furfurylaminopurine

NAA

Naphthalene acetic acid

PAA

Phenylacetic acid

Introduction

The production of doubled haploid plants has become a key tool in advanced plant breeding. Plant breeders are increasingly using this system in their mainstream programs in a large number of major crops (Kasha and Maluszynski 2003). In hexaploid bread wheat (Triticum aestivum L.), since Ouyang et al. (1973) first obtained green plants from anther cultures, protocol improvements have been made (Tuvesson et al. 2000; Barnabás et al. 2001; Zheng et al. 2001). Efficiencies have increased to a level adequate for the application of these techniques in commercial breeding programs. However, important differences exist within a given genus.

In tetraploid durum wheat (Triticum turgidum ssp. Durum (Desf.) Husn), several authors have reported its recalcitrant nature to anther culture. Green plants have been regenerated at a low frequency, because most of the genotypes did not respond at all to anther culture, or because most plantlets regenerated were albino (Hadwiger and Heberle-Bors 1986; Foroughi-Wher and Zeller 1990; Cattaneo et al. 1991; Orlov et al. 1993; Ghaemi et al. 1995; Otani and Shimada 1994; J'Aiti et al. 1999, 2000; Silva-Romano and Canhoto 1999; Dogramaci-Altuntepe et al. 2001; Jauhar 2003b).

In the study of Hadwiger and Heberle-Bors (1986), two green plantlets were obtained from anther culture, while in that of Ghaemi et al. (1993) three green and two albino plantlets were regenerated. Many genotypes did not respond at all to anther culture, as only one green and three albino plants were obtained from 15 different cultivars by J'Aiti et al. 1999, 2000). The best results were reported recently by Dogramaci-Altuntepe et al. (2001) when they produced, using ten Turkish cultivars, 248 green plants from 86,400 anthers. However, the chromosome number in the green plantlets ranged from 28 to 70 (only 52 plants being tetraploids). Genotype interactions with stress pre-treatments (Mentewab and Sarrafi 1997), medium composition (Ghaemi et al. 1994) and cytoplasmic types (Ghaemi et al. 1993) were described.

Therefore, two important problems remain to be solved in the durum wheat anther culture technique. The first, the high proportion of albino plants regenerated. For a great number of genotypes, most of the plantlets regenerated are albino. The second is the varying interactions between genotype × growth condition and genotype × induction medium (Jauhar 2003b). These two problems can make the anther culture system difficult to use for breeding programs. For these reasons, until now, the intergeneric crossing with Zea mays L. has been the method used to produce haploid plantlets (Jauhar 2003a).

The aim of this study was to develop a protocol to produce doubled haploid durum wheat plants that could be applied to most genotypes. Many experiments were made to select media and other conditions for the protocol and then it was applied to a number of genotypes. Results on spontaneous chromosome doubling are reported and the various factors influencing success are discussed.

Materials and methods

Seeds of donor plants were sown in a paper-pot with an equal mixture of peat, vermiculite and sand. Plants were vernalized for 1 month in a growth chamber at 4°C temperature, with 8/16 h light/dark photoperiod. Plantlets were transplanted to pots (15 cm diameter) with the soil mixture described above and kept in a growth chamber for 1 month at 12°C with a 12 h photoperiod. After the first month, the temperature was increased to 18–21°C and the photoperiod lengthened to 16 h light. Relative humidity was kept at 70–80%. An N: P: K (15:15:15) fertilizer was mixed into the soil and a foliar fertilizer was applied once a week.

Spikes were collected when most of the microspores were at the mid- to late-uninucleate stage. Collected spikes were removed from their sheaths in a flow bench and sprayed with 70% ethanol. The three anthers of 10 flowers (five from each side) from the central part of each spike were plated on a pre-treatment medium containing 0.7 M mannitol and 30 mM CaCl2.2H20 solidified with 8 g/l agarose. In each case, 30 anthers were used from each spike. Plates were kept for 5–6 days at 24°C in the dark. Microspore isolation was performed according to the barley protocol used by Cistué et al. (1995) including the use of one homogenizer to macerate the anthers, and the separation of viable microspores by centrifugation on 20% maltose band. While six microspore isolations were performed for the experiment 1 and three more for the experiment 2, a total of 35 microspores isolations were carried out to examine the response of genotypes using the protocol developed. While many experiments were sequentially performed to develop the protocol used, only the last two which summarize the current state of the technique will be described here.

The microspores separated from anthers were suspended in 0.3 M mannitol. Viable microspores were inoculated on 1.5 ml of C17 medium (Wang and Chen 1986) in 3 cm plates with 300 g/l of Ficoll-400 and ten ovaries per plate obtained from another spike, at a more advanced stage of development, from the same genotype. The auxins used depended on the experiment.

Experiment 1

The induction medium C17 with 300 g/l Ficoll-400 was used with two different hormone compositions. The medium A had 1 mg/l 2,4-D plus 1 mg/l BAP. The medium B had similar amount of auxin but, instead of BAP, 1 mg/l KIN was added to the medium. The Italian cultivar Simeto and the Spanish cultivar Arcolino (Batlle Seeds S.A.) were used, employing three replications for genotype. For each microspore isolation 720 anthers were extracted from 24 spikes and cultured. The 72 spikes belonged to the same batch of plants, growing at the same time in the same growth chamber. Analysis of variance was carried out using the GLM procedure of SAS.

Experiment 2

For this experiment the induction medium composition was the same as in experiment 1. The mature embryoids obtained from three microspore isolations, of the cultivar Antón and the selected line L-215 (Batlle Seeds S.A.), were transplanted on to two different regeneration media. From each microspore culture 60 embryoids were extracted (30 for each regeneration medium). The medium A was J25-8 as used by Jensen (1983), while medium B was the modified FHG medium used in our laboratory for culturing barley embryoids (Castillo et al. 2000). The plates were kept for 15 days at 24°C with a 16 h photoperiod. Number of regenerated embryoids and number of green and albino plantlets were counted. Analysis of variance was carried out using the GLM procedure of SAS.
Table 1

Comparisons between KIN and BAP in C17 induction media for durum wheat, based upon two genotypes and three replications. Numbers are average of three replications. Embryoids plus structures are equal to divisions. Twenty-four spikes (720 anthers) were used for each microspore isolation and genotype (half for each treatment)

Genotype

KIN/BAP

Anth./Isol.

Divisions

Embryoids

Structures

Albino. plants

Green plants

Arcolino

KIN

360

54.3b

20.3b

34.0b

5.7a

10.7b

Arcolino

BAP

360

136.0a

56.7a

79.3a

7.0a

42.7a

Simeto

KIN

360

63.0b

30.3b

32.7b

7.3b

5.7b

Simeto

BAP

360

103.3a

53.7a

49.7a

20.0a

13.7a

Means followed by different letters are significantly different at p=0.05 level

For the evaluation of the new protocol on different genotypes, a wide range of genotypes, including three pure lines and 20 different F1 crosses provided by private companies who have partially supported this activity were used. The cultivars Antón, Durtres and Semolón were used as reference genotypes as they are extensively grown throughout Spain. The origin of two of the crosses was known, Peñafiel by Regallo and Regallo by Peñafiel (Agromonegros Seeds S.A.), whereas the other 18 F1 represented a wide collection of crosses of unknown pedigree to us. However, according to Batlle Seeds S.A., the origin of their parents was genotypes from France, Mexico (CIMMYT) and Spain. They were chosen by the company breeders based just on their potential breeding value and not for their capacity to give haploid plants, which was unknown.

Based on the results from the two previous experiments and others not reported, viable microspores were inoculated on 1.5 ml of C17 medium in 3 cm plates with 300 g/l of Ficoll-400 and 10 ovaries per plate. The medium was supplemented with 1 mg/l 2,4-D and 1 mg/l BAP. The plates were kept for 20 days at 24°C in the dark. After that period another 1.5 ml of fresh medium was added. A total of 35 microspore isolations were cultured following this protocol.

After 25–30 days embryoids and less organized structures were transferred to J25-8 medium (Jensen 1983) in 5 cm plates solidified with 3 g/l Gelrite. The plates were kept for 15 days at 24°C with a 16 h photoperiod. Green plantlets were transferred to Magenta boxes with the Murashige and Skoog (1962) medium salts at half strength, and supplemented with 1 mg/l NAA.

These boxes were placed in a vernalization growth chamber at 4°C with 8 h light photoperiod. After 1 month plantlets were transferred to a well-aerated soil in a growth chamber with 90% humidity and 12 h photoperiod. After 15 days, plants were transferred to the greenhouse until seed was harvested. At the three to four leaf stage the ploidy level, of 257 green plantlets from just six high-responding crosses, was studied with a Partec flow cytometer. Spontaneously doubled plants were directly potted into soil, while haploid plantlets were subsequently treated with colchicine using the system of Thiebaut et al. (1979). The plants obtained from all genotypes and crosses were grown to maturity in order to quantify the seed production in the spikes.

Results

Experiment 1 showed that both BAP and KIN in induction media result in plant regeneration from durum wheat microspores. However, BAP significantly improved embryoid formation and green plant regeneration compared to KIN. Besides, BAP seems to produce embryoids with two or three shoots very frequently. This is very useful to regenerate plants with numerous spikes. The response difference between the two genotypes was also significant (Table 1).

The embryoids from the experiment 2 regenerated significantly better on the J25-8 medium than on the FHG modified regeneration medium normally used for barley embryo regeneration. From 180 embryos plated for each of the two genotypes, 33 (5.7 + 5.3) × 3 regenerated into green plants in the J25-8 medium while only 14 (2.3 + 2.3) × 3 regenerated in the FHG medium (Table 2). This was consistent across the two genotypes.
Table 2

Comparison between two regeneration media FHG and J25-8 on durum wheat embryoids. Numbers are average of three replications. Three microspores isolations per genotype, extracting from each one 60 embryoids, and inoculating 30 per regeneration medium

Genotype

Medium

Embryoids

Regenerated plant

Albino plant

Green plant

Antón

FHG

30

4.0b

1.7b

2.3b

Antón

J25-8

30

11.0a

5.7a

5.3a

L-215

FHG

30

8.3b

6.0a

2.3b

L-215

J25-8

30

15.3a

9.7a

5.7a

Means followed by different letters are significantly different at p = 0.05 level

For the study of the survey of genotype responses, the results obtained from 35 microspore isolations are shown in Table 3. It is important to note that all crosses gave green plants, even though the efficiency of green plants regenerated appeared to be genotype-dependent. While statistics could not be run on such a survey, it also appears that the number of dividing microspores, the proportion between embryoids versus structures obtained, and the proportion between green versus albino plants regenerated were also genotype-dependent. Four hundred and seven green plantlets were obtained from 35 microspore isolations (16,050 anthers). As an average 2.5 green plants were obtained per 100 anthers. From the 407 green plantlets obtained, 275 were completely fertile, without any type of abnormality in the morphology neither of the plant nor of the spike. The others were haploid or sterile.
Table 3

A survey of the number of embryoids, less organized structures and green and albino plants produced in 35 microspore cultures of various durum wheat genotypes using the protocol described in this study

Mic. isolat.

Genotype

Anthers/per plate*

Embryoids

Structures

Albino plants

Green plants

Fertile plants

1

Antón

180

31

400

2

2

2

2

Antón

360

18

200

8

4

3

3

Antón

840

22

952

0

2

2

4

Durtres

270

15

470

0

2

2

5

Semolon

270

21

2

6

1

1

6

Agr. Peñafiel × Regallo

450

97

6

46

4

1

7

Agr. Peñafiel × Regallo

450

24

0

2

11

3

8

Agr. Peñafiel × Regallo

480

32

0

3

9

5

9

Agr. Peñafiel × Regallo

600

10

0

0

5

3

10

Agr. Peñafiel × Regallo

480

47

0

2

13

6

11

Agr. Peñafiel × Regallo

720

64

0

9

38

28

12

Agr. Regallo × Peñafiel

300

14

0

3

3

2

13

Agr. Regallo × Peñafiel

810

12

0

1

5

5

14

Agromonegros Cl × As

300

12

50

2

1

0

15

Batlle 1×18

300

21

10

0

4

2

16

Batlle 2×18

570

7

150

0

2

0

17

Batlle 2×19

480

3

0

2

1

0

18

Batlle 3×18

480

131

175

56

14

8

19

Batlle 3×19

390

14

12

1

1

0

20

Batlle 5×18

360

51

50

3

34

30

21

Batlle 5×18

300

110

165

0

89

62

22

Batlle 5×18

450

73

185

10

62

48

23

Batlle 5×19

420

27

30

7

7

4

24

Batlle 6×19

240

17

53

5

6

3

25

Batlle 10×19

690

6

6

0

6

3

26

Batlle 10×19

480

70

30

4

32

21

27

Batlle A88×C30

600

434

285

114

20

16

28

Batlle B97×C30

420

49

109

22

3

1

29

Batlle B97×S88

420

70

780

34

3

0

30

Batlle DP×C30

420

174

72

76

1

1

31

Batlle E13×C30

600

165

26

65

4

1

32

Batlle E30×C30

330

36

0

4

2

1

33

Batlle E42×C30

510

51

0

23

6

3

34

Batlle SV×C30

510

108

32

28

2

1

35

Batlle SV×C30

570

59

0

5

8

7

 

Total

16,050

2,095

4,250

543

407

275

*The number of anthers extracted from one spike was always 30.

The developmental stage of both the microspores (Fig. 1a) and the ovaries used for co-culture greatly influenced the frequency of divisions. Following this protocol, mature embryoids were obtained in all isolations (Fig. 1b and c), which frequently regenerated into healthy green plantlets (Fig. 1d). It was fairly common to obtain embryoids with two or three shoots (Fig. 1e). In many cases, vigorous green plantlets were regenerated (Fig. 1f).
https://static-content.springer.com/image/art%3A10.1007%2Fs00299-005-0047-8/MediaObjects/299_2005_47_Fig1_HTML.gif
Fig. 1

Different stages in the development of the microspores in culture of durum wheat. (a) Microspores in the uninucleate vacuolate stage. (b) First embryogenic divisions seen after 20 days on culture. (c) Mature embryoid (scutellum, coleoptile and coleorhiza) produced after 30 days on induction medium. (d) Embryoid regenerating green plantlet. (e) Embryoid regenerating with two green shoots. (f) Regeneration plate 10 days after transplanting. Three green plantlets and one plantlet without chlorophyll

The results of ploidy level analysis in 257 green plantlets obtained from six high responding crosses are summarized in Table 4. Spontaneously doubled haploids accounted 67% of plants, whereas 28% were haploids and a very few had other numeric chromosomal constitution. These doubling frequencies appear to be influenced by the genotype to a certain degree. From the 166 doubled haploid plants, 149 were completely fertile. Therefore, using this protocol, the aneuploidy if any, was not very high. Batlle Seeds Company has already selected two doubled haploid lines produced from these microspore cultures for potential variety release, based on their high grain yield assessed in two years of field trials.
Table 4

Ploidy levels of 257 green plantlets produced by microspore culture in six crosses of durum wheat, checked at the three to four leave stage by flow-cytometry. Number and percentage of fertile plants

F1 crossing

2× (H)

4× (DH)

Number of plants

Fertile plants DH

% Fertile

Batlle 3×18

0

1

10

0

0

11

7

70

Batlle 5×18

46

0

114

1

7

168

99

91

Batlle 5×19

4

0

3

0

0

7

3

100

Batlle 6×19

2

0

2

0

1

5

2

100

Batlle 10×19

20

0

17

0

0

37

14

88

Peñafiel × Regallo

1

0

26

0

2

29

24

92

Number of Plants

73

1

172

1

10

257

149

 

%

28.4

0.4

66.9

0.4

3.9

  

90

Discussion

Every single step in the technique significantly influenced production of mature embryoids. The protocol for growing the microspore donor plants has already been suggested as a key factor for the success of the microspore culture technique (Zheng 2003). The growing of vigorous microspore donor plants increased the number of green plants regenerated. Durum wheat is very demanding in fertilization and in an exact control of humidity, photoperiod and temperature, much more than barley. In order to obtain this kind of plant, it was very important to carefully monitor irrigation and fertilization. Durum wheat usually produces fewer tillers than barley or bread wheat. For that reason it is very important to have optimum plant conditions. Fertilization during the second month of cultivation was one of the most important factors to obtain healthy donor plants. Environmental stresses both biotic and abiotic should be minimized. In this study, we have used microspores donor plants grown in an environment controlled chamber as field-grown spikes frequently had fewer healthy microspores at the right culture stage. In every case where we made microspore isolations using spikes collected in the field or from plants growing with low fertilization, all the microspores died without beginning divisions.

Anthers pre-treatment also was one of the most important steps. Most of the authors preferred to maintain the spikes in cold temperature at 4°C during a week or longer periods or high temperature at 33°C for 2–3 days (Zheng 2003). Other pre-treatments have also been used such as starvation, osmotic shock, or microtubule disruption agents (Jähne and Lörz 1995). Sugar starvation during 4–5 days using mannitol instead of sucrose or maltose is commonly used in barley (Robert-Oehlschlager and Dunwell 1990; Cistué et al. 1994, 1999). With this system, the number of embryoids increased and as a consequence the number of green plants regenerated also increased. Besides, according to Cistué et al. (1995), in barley albinism increased proportionately to the length of time on induction medium. This is very important for durum wheat microspore culture because the high albinism frequency is one of the most serious problems as previously mentioned.

Microspore isolation can be done by blending, stirring, maceration or anther shedding. According to Gustafson et al. (1995), blending maximizes microspore viability up to 75%. However, we used maceration because the anthers were already isolated as a consequence of the pre-treatment in mannitol.

We also found that the addition of 5–10 ovaries for every 3 cm plate was critical for success. These ovaries must be in a healthy condition and removed at a more mature stage of the spike than for anther collection, such as when the microspores are in the tri-nucleate stage (Castillo and Cistué 1993). This co-culture has already proved to be very important in other species such as barley or bread wheat (Hu and Kasha 1997; Devaux and Li 2001; Zheng et al. 2002). When ovaries were added to durum wheat microspore cultures, the number of embryoids increased significantly.

Hormones also play an important role in both induction and regeneration. In Table 1, the use of BAP significantly improved induction when compared to equal amounts of KIN. The numbers of microspores dividing was highly genotype-dependent. This aspect seemed to be associated with the growing conditions of the microspore donor plants and with the induction media used. Anthers from different genotypes can have different hormone concentrations (Cho 1991) and, thus, the concentration of hormones used in the induction medium may not be at the optimum for all genotypes. However, according to our experience, a high number of divisions was not necessarily correlated with a high number of green plants regenerated, because some structures degenerate into callus.

Mature embryoids (scutellum, coleoptile and coleorhiza) were obtained. Other authors have used PAA in the induction medium (Hu et al. 1995) or a combination of PAA with 2,4-D and kinetin (Zheng et al. 2001). We used 2,4-D and BAP after several preliminary experiments (data not shown) as this combination seemed to be the best. Another key aspect in the success of the technique was the medium osmolarity. Kang et al. (2003) already observed in spring wheat that increased osmolarity within certain ranges resulted in a larger number of green plants regenerated. In our technique Ficoll was used for osmolarity control. Ficoll was also useful in triticale anther culture (Immonen and Robinson 2000).

Ficoll is often considered too expensive to be used for the routine production of doubled haploids. However, as we were using only 3 ml per plate, because a high concentration of microspores was necessary, only 105 ml of C17 Ficoll medium was used for the 35 isolations. These figures translate in to the use of 31.5 g of Ficoll to obtain 407 green wheat durum haploid plants, most of which were doubled; thus, we believe it does not represent a too high cost.

In our laboratory, we use the same protocol for bread wheat and durum wheat, with the exception of employing the MS-modified medium (Kasha et al. 2003) instead of C17 for bread wheat. Use of the same protocol saves time and material when working with different species.

The proportion of green plants versus albino was 43% (407/950). This result was comparatively good according to previous studies with durum wheat, where albinism seemed to be the major problem (Saidi et al. 1997). The high proportion of regenerated green plants was probably a consequence of several aspects: the genotypes used, the stage of the spike, the pre-treatment of the anthers, the production of mature embryoids avoiding callus production and the regeneration medium used. In this work, 407 green plantlets were produced, using 35 microspore isolations, which means an average of 11.6 green plants per plate.

One example of the genotype influence was the cross 5 × 18 where from three isolations with 1,110 anthers used, we obtained 13 albino and 185 green plants. On the contrary with Regallo × Peñafiel, two microspore isolations and also 1,110 anthers used, we only obtained four albino and eight green plants.

We utilized the J25-8 medium (Jensen 1983) because it is an optimum medium for embryoid germination. It was originally developed to maximize embryo regeneration in the interspecific hybridization of H. vulgare and H. bulbosum, crosses used to produce haploid barley plants. Embryoids regenerated from durum wheat usually grew slowly but a great number of plants were regenerated in this medium.

The reason for the high spontaneous chromosome doubling among regenerated plants is unknown. Kasha et al. (2001) and Shim and Kasha (2003) proposed that the mannitol pre-treatment increased nuclear fusion following the first mitotic division in the microspore. However, more studies should be undertaken in this aspect for durum wheat.

Until now the system used to obtain durum wheat haploid plants has been the intergeneric crossing with Z. mays L. (Jauhar 2003a). This system has the advantage that all the haploids obtained are green. For some genotypes, per 100 flowers pollinated is possible to obtain 4,7 haploid green plantlets. However, this system has two problems. Firstly the problem of growing to flowering corn and durum wheat, at the same period of the year. Secondly, the use of colchicine is compulsory and this is a carcinogenic substance. Therefore, the use of microspore culture, if the numbers of plantlets produced are similar to the ones of intergeneric crossing, could be a good alternative.

In order to improve this protocol, more research is currently underway on different aspects. For example, we are assessing the use of chemical inducers formulations before the pre-treatment because they worked very well with T. aestivum (Zheng et al. 2001). Furthermore, we are studying the influence of arabinogalactans (AGP) instead of ovaries in the induction medium (Kasha et al. 2003). On the other hand, the genetic basis of the response to microspore culture needs to be investigated in a wide collection of durum wheat germplasm to determine the extent of genotypic dependence of the protocol. But at present, this protocol works well for several F1 crosses like Peñafiel × Regallo or Batlle 5 × 18, where 38 and 89 green plants per 720 and 300 anthers were obtained, respectively. These results are similar or even better than the results obtained for the integeneric crossing with maize, and obtaining spontaneous doubled haploids. Besides, with this protocol is possible to obtain complete embryos which develop normal green doubled haploids plants. This is essential to perform new experiments to improve the protocol. Thus, our procedures have the potential to be implemented in applied breeding programs.

Acknowledgements

We would like to thank Prof. K. J. Kasha (University of Guelph, Canada) for the extensive revision of the manuscript and by his useful suggestions. We also want to thank Prof. I. Romagosa (University of Lérida, Spain) for helping us with the statistical analysis. We also want to thank Batlle Seeds S.A. and Agromonegros Seeds S.A. for their support of this research. The research was also supported by Projects 2FD1997-1258, PTR1995-0732-OP and AGL2002-04139-C02 from the Spanish Ministry of Science and Technology.

Copyright information

© Springer-Verlag 2005