Mature embryo axis-based high frequency somatic embryogenesis and plant regeneration from multiple cultivars of barley (Hordeum vulgare L.)

Abstract

A highly reproducible regeneration system through somatic embryogenesis from the excised mature embryos (MEs) of dry seeds of a range of European barley cultivars was developed. By minimizing the germination of plated MEs, primary callus could be obtained with high frequency which permitted efficient embryogenesis and regeneration of a large number of green plants. Different approaches were tested to reduce or prevent normal germination: (i) the use of a well defined balance of maltose and 2,4-D in the induction medium, (ii) soaking of seeds in water containing 2,4-D solution, (iii) direct culture of excised embryonic axes, (iv) longitudinally bisected MEs giving two halves, and (v) complete removal of the elongated main shoot including any roots within a week of culture initiation. Culturing of bisected MEs and whole embryonic axes gave the best responses with respect to large amounts of callus combined with minimal germination. The incorporation of BAP at low levels in the medium was found to be most effective for embryogenesis and the maintenance of long-term morphogenic capacity (more than 11 months up to now). This procedure allows the complete regeneration of plants in 16–20 weeks, from the initial isolation of MEs through all the steps to the development of plants ready to be transferred to the soil. The protocol was first developed for cv. Golden Promise and successfully applied to commercial cultivars. All cultivars tested formed embryogenic callus, with overall rates ranging from 22–55% and an average number of green plants per embryogenic callus from 1.5 to 7.5 across the genotypes.

Barley, commercial cultivars, embryogenesis, embryonic axis, mature embryo, primary callus, plant regeneration

Introduction

All biotechnological approaches like genetic engineering, haploid induction, or somaclonal variation to improve traits of important crops strongly depend on an efficient recovery of plants through in vitro systems. Previous studies had determined that immature cells and tissues are the best type of explants for plant regeneration, especially in recalcitrant crops like monocotyledonous species. For all the important cereals, efficient and reproducible plant regeneration protocols have been developed, mostly based on immature embryos (Vasil, 1994; Repellin et al., 2001). Consequently, transgenic plants could be generated from all the major cereals by gene transfer into the scutellum of immature embryos or into embryogenic callus derived from these explants (Gordon-Kamm et al., 1990; Christou et al., 1991; Vasil et al., 1992; Wan and Lemaux, 1994). However, the successful application of these methods is determined by the genotype (Lührs and Lörz, 1987; Popelka and Altpeter, 2001), donor plant quality (Maës et al., 1996; Dahleen, 1999), developmental stage of the explant (Thomas and Scott, 1985; Maës et al., 1996), and composition of the culture medium (Lührs and Lörz, 1987; Barro et al., 1999; Dahleen and Bregitzer, 2002) resulting in inefficient in vitro culture protocols.

Recently, plant regeneration and subsequent generation of transgenic plants have been obtained in different cereals using readily available explants such as mature embryos (Torbert et al., 1998), mature seeds (Abedinia et al., 1997; Cho et al., 2004), leaf base segments (Gless et al., 1998; Chugh and Khurana, 2003), shoot meristems (Zhong et al., 1996; Goldman et al., 2003), or in vitro shoot meristematic cultures (Zhang et al., 1999). The use of mature embryos (MEs) has remarkable advantages compared with immature tissues as explants. Thus, the need for growing donor material in greenhouses under controlled environmental conditions requiring intensive labour, time, and space can be avoided, and especially for winter cultivars, no additional time input due to vernalization is required. Moreover, dry seeds are available in large quantity and year round with no problems due to seasonal influence on tissue culture response (Dahleen, 1999).

For barley, the suitability of MEs for callus induction (Bayliss and Dunn, 1979) and plant regeneration via organogenesis (Lupotto, 1984; Ukai and Nishimura, 1987) and somatic embryogenesis (Rengel, 1987) were described in previous reports, but with low regeneration frequencies. In recent time, MEs as explants have been re-evaluated and protocols with improved regeneration efficiency but with high cultivar dependency were provided (Akula et al., 1999; Ganeshan et al., 2003).

The establishment of a highly efficient plant regeneration system by somatic embryogenesis from the embryonic axes of mature barley embryos from eight cultivars with low genotype dependency, and the long-term maintenance of embryogenic lines, is reported here. The present protocol of efficient somatic embryogenesis and plant regeneration based on mature embryos could be used for genetic transformation of commercial cultivars of barley as another choice of regeneration system.

Materials and methods

Plant material

Mature dry seeds of different barley (Hordeum vulgare L.) cultivars given in Table 1 were used as source of mature embryos (MEs). The model tissue culture cv. Golden Promise was used for all the initial experiments as described below.

Table 1.

Characteristics of the cultivars of barley used in the present study

Cultivar	Plant breedera	Form	Spike type	Registration date	Seed production acreage (ha) in 2004 in Germanyb
Golden Promisecd	Miln/K.Lyn	Spring	Two rowed	1966	–
Salome	Langenstein	Spring	Two rowed	1981	–
Borwina	Gülzow	Winter	Six rowed	1982	–
Duet	Nickerson	Winter	Two rowed	1995	1457
Franziska	Secobra	Winter	Six rowed	2000	5393
Lomerit	v. Lochow	Winter	Six rowed	2001	5055
Masto	Semundo	Winter	Six rowed	1988	–
Merlot	Nordsaat	Winter	Six rowed	2002	3216

a

http://www.stmlf-design2.bayern.de/lbp/forsch/pz/gerstenstamm

b

Data were provided from Bundessortenamt, Hannover, Germany.

c

Cultivar with high tissue culture ability.

d

Still grown in the UK for brewing.

Isolation and culture mode of mature embryos and its sections

Seeds were de-husked, surface-sterilized as described by Müller et al. (1989) and soaked in sterile distilled water or in water containing 3, 6, or 10 mg l⁻¹ 2,4-D solution for 2 d at 4 °C. Thereafter, seeds were rinsed 3–5 times with sterile water and MEs were dissected from the endosperm after slightly damaging the radicle portion. Twelve MEs were placed embryonic axis-side down in full contact with 25–30 ml of primary callus induction medium in a Petri dish (92×16 mm). In parallel, the sections of the MEs from cv. Golden Promise such as scutellar tissues and the embryonic axes were excised and cultured separately to determine their callus induction response. Scutella were plated abaxial face-up while embryonic axes were cultured horizontally. Furthermore, MEs, longitudinally bisected into two halves, were also plated as for whole MEs.

Culture media for induction of primary callus

Nine different media were formulated for optimal primary callus induction with minimal germination of MEs from cv. Golden Promise. The basal medium used was the same as reported previously (Sharma et al., 2004) consisting of MS salts and vitamins (Murashige and Skoog, 1962), supplemented with 3, 6, or 9% (w/v) maltose corresponding to an osmolality of 180, 270, and 360 mOsmol kg⁻¹ H₂O, respectively, and 3, 6, or 10 mg l⁻¹ 2,4-D (Sigma, Taufkirchen, Germany). These media formulations were referred as: 3M3D (3% maltose+3 mg l⁻¹ 2,4-D), 3M6D (3% maltose+6 mg l⁻¹ 2,4-D), 3M10D (3% maltose+10 mg l⁻¹ 2,4-D), 6M3D (6% maltose+3 mg l⁻¹ 2,4-D), 6M6D (6% maltose+6 mg l⁻¹ 2,4-D), 6M10D (6% maltose+10 mg l⁻¹ 2,4-D), 9M3D (9% maltose+3 mg l⁻¹ 2,4-D), 9M6D (9% maltose+6 mg l⁻¹ 2,4-D), and 9M10D (9% maltose+10 mg l⁻¹ 2,4-D). To observe the effect of other auxins on callus induction media 6M3D, 6M6D, and 6M10D, 2,4-D was substituted by equal amounts of Picloram or Dicamba. The influence of the cytokinins BAP or TDZ in combination with 6 mg l⁻¹ 2,4-D in the basal medium containing 6% maltose was also investigated for Golden Promise.

All media components except Dicamba were mixed together, adjusted to pH 5.8, and solidified with 0.3% (w/v) Gelrite prior to autoclaving at 121 °C for 20 min. Filter-sterilized Dicamba was added to the media after autoclaving. Unless indicated otherwise, the chemicals used in this report were purchased from Duchefa (Haarlem, The Netherlands).

Primary callus induction

The isolated MEs were cultured on all nine media described, incubated in the dark at 25 °C for 4–5 weeks, and observed weekly under a stereo microscope. After 5–7 d, embryos started callus formation and germination, shoots and roots were cut out completely, and the callus responding explants plated again in the same Petri dish for callus proliferation. Further elongated main shoot and/or roots, if any, were removed and the non-responding scutellar tissues of the intact callus were separated and discarded after 2–3 weeks. Five weeks later, callus induction quality and type was evaluated as described (Ryschka et al., 1991). The amount of callus was scored visually as low (+; 50–100 mg), moderate (++; 150–200 mg), or high (+++; 250 mg or more). Callus derived from each ME was maintained separately until the completion of plant regeneration.

Embryogenic callus formation and maintenance

Primary callus with or without embryoids was divided into 2–4 pieces (50–100 mg, depending on size and amount) and transferred to embryogenic callus induction medium 3M3DB (3% maltose+3 mg l⁻¹ 2,4-D+0.01 mg l⁻¹ BAP, Sigma) or 3M6DB (3% maltose+6 mg l⁻¹ 2,4-D+0.01 mg l⁻¹ BAP). Cultures for embryogenesis were incubated for 1 week in the dark at 25 °C, followed by 2–3 weeks under low light conditions (20–30 μE m⁻² s⁻¹) at 25 °C and a 16/8 h (light/dark) photoperiod. The amount of embryogenic callus was scored similarly as above for primary callus. Each week, embryogenicity was monitored stereomicroscopically and callus with somatic embryoids was subcultured on 3M3DB or 3M6DB medium for proliferation and subsequent maintenance by biweekly subcultures. Medium containing 2,4-D as above in combination with TDZ (0.01 mg l⁻¹) instead of BAP was also used. To determine the effect of other auxins on the maintenance of embryogenic callus from Golden Promise, basal medium supplemented with 3% maltose+3 or 6 mg l⁻¹ Picloram or Dicamba+0.01 mg l⁻¹ BAP or TDZ was tested.

Plant regeneration

After 5–6 weeks with one subculture, embryogenic callus with or without green structures and/or shoot primordia was transferred to medium containing 3% maltose+1 mg l⁻¹ 2,4-D+0.1 mg l⁻¹ BAP (referred to as 3M1DB) for regeneration and incubated for 2–3 weeks under low light conditions as above. At the end of 3 weeks, most of the embryogenic callus turned into shoots/plantlets. All regenerated green and albino shoots/plantlets were counted from each callus line, and callus that did not regenerate completely was subcultured again on 3M1DB medium for a further 2–3 weeks. Finally, the number of shoots and plantlets was counted again and summed with the first counts to give the total regeneration. Other media were also used, where BAP was substituted by TDZ.

Rooting of regenerated shoots/plantlets

For the induction of a strong root system, well-developed shoots/plantlets (3–4 cm in length) from the regenerating somatic embryoids were transferred to rooting medium consisting of basal medium+3% maltose+0.5 mg l⁻¹ IBA devoid of casein hydrolysate, L-proline, and thiamine hydrochloride in glass vessels (diameter: 5.5 cm; height: 9.5 cm) covered with plastic lids (Magenta® B-cap, Magenta Corp., Chicago, USA). Cultures for rooting were exposed to a high light intensity (60–80 μE m⁻² s⁻¹) at 25 °C and a 16/8 h (light/dark) photoperiod for 2–3 weeks. To examine normal growth and fertility, well-rooted plants, 10 per genotype, with a similar number of control plants were transplanted to soil and grown under greenhouse conditions. Plants from winter-type cultivars were first vernalized (Sharma et al., 2004) and then kept in the greenhouse. Seeds were harvested at maturity from all plants after 5–7 months.

Establishment of the protocol for plant regeneration in other cultivars

Factors which markedly influence embryogenesis and regeneration, such as maltose concentration, type and concentration of auxins as well as cytokinins, were optimized with MEs from Golden Promise in order to develop and standardize the present protocol. Moreover, approaches were analysed to reduce germination by means of the culture mode of MEs. The most suitable parameters were further applied to MEs from other genotypes including agronomically important cultivars.

Experimental plan and statistical analysis

Three identical and independent experiments were performed for each cultivar. For cv. Golden Promise four repetitions were done. In each replicate 60–85 MEs were used. The data presented here are the average of these experiments. ANOVA was performed to test for differences in embryogenic callus induction and plant regeneration among the eight cultivars. The significance of group differences was determined according to BONFERRONI.

Results

Mechanical possibilities to reduce the germination of cultured mature embryos

Initial experiments with cv. Golden Promise using medium 6M6D showed that germination of MEs strongly reduces callus formation. After 3 d, the cultured MEs swelled, enlarged in size with the elongation of the radicle and plumule and, after 5–7 d, callus formation along with germination was observed. Slightly damaging of the radicle portion hampered root formation in the germinating MEs thereby reducing the elongation of the main shoot. Bisecting the MEs into two halves favoured more callusing with quite a reduced elongated main shoot (less than 1 cm) and almost no rooting. Furthermore, the mode of plating the MEs onto medium strongly influenced the developmental process: plating the whole or bisected MEs with the embryonic axis-side down inhibited root elongation and facilitated callus initiation starting from day zero. Explants plated with the embryonic axis-side up with or without full contact onto the medium showed 100% germination. These germinating embryos mostly lifted off of the medium with either root or shoot side and gave no callusing response on any of the media tested.

Effect of 2,4-D and maltose on callus induction from whole mature embryos

The effect of nine different media for optimal primary callus induction with minimal germination of the ME explants of cv. Golden Promise was compared. There was a clear correlation between maltose and 2,4-D levels on the germination and callusing response (Fig. 1). Higher concentrations of maltose and 2,4-D reduced the main shoot and root elongation by up to 1–2 cm and simultaneously promoted primary callus induction with an optimum at 6% maltose combined with 6 or 10 mg l⁻¹ 2,4-D as contained in media 6M6D or 6M10D. Media with 9% maltose also strongly inhibited normal germination but callus induction was found to be low independent of the amount of 2,4-D. Despite a reduction of germination due to mechanical and chemical treatment, callus proliferation was further promoted by removing shoots and roots after 1 week of culture initiation from whole as well as bisected MEs which were plated again on the same medium for a further 3–4 weeks.

Fig. 1.

Influence of maltose and 2,4-D on primary callus induction and germination from mature embryos of barley cv. Golden Promise. Germination intensity was scored after 1 week as: + least germination (only main shoot elongation, 1–3 cm in length, with no or reduced roots); ++ moderate germination (main shoot elongation 3–5 cm in length with 1–2 cm long roots); +++ prominent germination (main shoot 5–10 cm in length with roots up to 3 cm). The amount of callus was scored as given in the Materials and methods. Sixty to 70 MEs were used as starting material.

Comparison of 2,4-D, Picloram and Dicamba on primary callus induction

Substitution of 2,4-D by equal amounts of Picloram or Dicamba in the callus induction medium containing 6% maltose did not result in any remarkable differences in germination response (data not shown) and subsequent initiation of primary callus and embryogenesis (Fig. 2). The amount and texture of the callus induced was similar on all three auxins tested after 3 weeks. However, after 5 weeks, 6 as well as 10 mg l⁻¹ of 2,4-D or Picloram evoked somatic embryoids on the surface of primary callus with the best response on 6 mg l⁻¹ of 2,4-D as in media 6M6D. Therefore, medium 6M6D was used as the initial callus induction medium in all subsequent experiments.

Fig. 2.

Influence of auxin type and concentration on the induction of primary callus and embryogenesis from mature embryos of barley cv. Golden Promise. Basal medium with 6% maltose was used. For the induction of embryogenesis all media contained additionally 0.01 mg l⁻¹ BAP. Average data from two independent experiments each with 60 MEs as the starting material.

Effect of cytokinins on primary callus induction

The addition of BAP or TDZ at levels higher than 0.1 mg l⁻¹ in the induction medium 6M6D promoted shoot organogenesis and/or multiple shoot formation rather than somatic embryogenesis. However, very low doses of BAP (0.001 mg l⁻¹) slightly enhanced embryogenesis whereas TDZ did not (Fig. 3).

Fig. 3.

The influence of BAP and TDZ added to medium 6M6D on primary callus induction from MEs of cv. Golden Promise after 4 weeks and on the frequency of the primary calli forming somatic embryoids after subculture on the same cytokinin concentration (medium 3M6D) recorded after another 4 weeks. Formation of multiple shoots/buds was also scored. Average data from two independent experiments each with 50 MEs as the starting material.

In vitro response of various tissues of mature embryos

The particular tissues of a ME such as the embryonic axis and scutellum were excised, cultured on 6M6D, and incubated for 4–5 weeks in the dark to identify the real source for callus origin. By culturing whole MEs, callus induction and somatic embryogenesis generally occurred from shoot meristematic and/or mesocotyl tissues of the embryonic axes, whereas scutella did not react with callusing (they only enlarged in size and later turned brown). These scutella inhibited callus proliferation, thus discarding them from the already developed callus within 2–3 weeks promoted primary callus induction (Fig. 4A). The plated embryonic axes showed no or low germination and evoked more callusing compared with the whole embryos after 4–5 weeks of incubation (Table 2). It was observed that, due to direct contact of the embryonic axis with the medium, a soft, yellowish cream and embryogenic type of callus developed from the plumule part after 1 week (Fig. 4B), while a soft, white and watery callus with small hairy roots was obtained from the radicle portion. Furthermore, it was also observed that the callus formed from the radicle grew faster than the callus of plumule origin, thus reducing the formation of embryogenic callus. Therefore, callus originating from the radicle together with any roots has to be completely removed within a week of culture initiation, and the remaining callus could be kept on the same medium for 4–5 weeks for further growth and embryogenesis. Isolated scutellar tissues shrunk, turned brown, and did not show any response even after an incubation period of 4–5 weeks (Fig. 4C). Despite these clear differences in the in vitro response, for practical reasons whole MEs were used in the following experiments since careful observation and removing of tissues which suppress embryogenic callus was necessary both for whole MEs as well as for isolated embryonic axes.

Fig. 4.

Primary callus induction, somatic embryogenesis, and plant regeneration from isolated mature embryos and embryonic axes of barley cv. Golden Promise. (A) Primary callus from whole mature embryos, (B) primary callus from embryonic axes, (C) non-responding mature scutellar tissues, (D) embryogenic callus, (E) as in (D) developing somatic embryos at late globular and heart-shaped stages, (F) maintenance of embryogenic callus in the Petri dish (92×16 mm), (G) as in (F) proliferating somatic embryos, (H) regeneration of shoots/plantlets in the Petri dish (92×16 mm), (I) as in (H) regenerating somatic embryos into small shoots and plantlets, (J) well-rooted shoots (6–8 cm), (K) fertile plant. Bars represent 10 mm, except for (E), (G), and (I) where the bar represents 4 mm.

Table 2.

Frequency of primary and embryogenic callus induction from different portions of mature embryos in cv. Golden Promise^a

Explant type	Number of explants used	Explants forming
			Primary callus (%)b	Embryogenic callus (%)c
Mature embryo	84	69 (++)d	52.4 (+++)
Embryonic axis	84	94 (+++)	66.7 (+++)
Scutellar tissue	84	0	0

a

Two independent experiments were carried out, average data.

b

Frequency of primary callus induction was evaluated by counting the explants forming callus out of the total number of explants cultured ×100.

c

Frequency of embryogenic callus formation was assessed by the percentage of explants forming embryogenic callus and the total explants used.

d

Amount of callus scored visually as described in the Materials and methods.

Preincubation of seeds in water containing 2,4-D

Sterilized seeds of Golden Promise were soaked in water containing 3, 6, or 10 mg l⁻¹ 2,4-D in order to examine the influence of soaking seeds in 2,4-D solutions on germination and callus formation from MEs after isolation and culture on induction medium 6M6D. Intact MEs in seeds showed swelling and enlargement similarly to seeds soaked in plain water. However, upon culturing, a lowered elongation of the main shoot and root by up to one-third with more callus formation was observed after preincubation in all the 2,4-D concentrations added in water. The 10 mg l⁻¹ treatment caused softening of the scutellar tissues in the intact seeds which made the isolation of MEs difficult. Although pretreatment in water plus 3 or 6 mg l⁻¹ 2,4-D resulted in a slightly higher primary callus induction (approximately 70% for 3 mg l⁻¹ and 75% for 6 mg l⁻¹) compared with 64% after incubation in plain water, no influence on the frequency of embryogenic callus induction was found. Seeds soaked in water containing 6 mg l⁻¹ 2,4-D were used in all the following experiments as the source of MEs.

Embryogenesis and callus maintenance

The basal media 3M3D and 3M6D were found to be optimal for the induction of embryogenic callus from already formed primary callus when incubated for 1 week in the dark followed by 2–3 weeks under low light. An increase of 2,4-D up to 10 mg l⁻¹ as in medium 3M10D favoured the formation of non-embryogenic callus. The addition of 0.01 mg l⁻¹ BAP (3M3DB or 3M6DB) promoted the efficiency of embryogenic callus proliferation and subsequent maturation of early formed embryoids from the globular stage to the heart-shaped stage (Fig. 4D, E). Replacement of BAP by TDZ did not promote embryogenesis, but stimulated the formation of shoot buds with the subsequent development of small shoots. Both the BAP-containing media 3M3DB and 3M6DB were also found to be suitable for maintaining selected highly embryogenic callus. The best reactivity was observed on 3M6DB with biweekly subculture with respect to longevity of regenerability (Fig. 4F, G). The successive subcultures gave rise to the formation of numerous clusters of embryoids at different developmental stages with leaf primordia and/or small shoots. Moreover, plantlets were also developed by precocious germination of already matured somatic embryos in the old cultures. Removal of small shoots/plantlets was necessary during every subculture to promote the proliferation of embryogenic callus. Passage on a medium devoid of BAP resulted in the loss of embryogenic potential.

Plant regeneration and rooting of shoots/plantlets

Spontaneous formation of shoots and plantlets from embryogenic callus was observed on 3M3DB or 3M6DB. Moreover, plantlet development could be induced by the transfer of embryogenic calli having heart- and torpedo-shaped embryoids to regeneration medium 3M1DB (Fig. 4H, I). Replacement of BAP by a similar level of TDZ also favoured regeneration (Table 3). Media free of BAP or TDZ or supplemented only with 2,4-D (1 mg l⁻¹) were not effective for improved differentiation (data not shown). Regenerated shoots and plantlets were removed from the main cultures and the formation of plants with strong root systems was achieved on rooting medium within 2–3 weeks (Fig. 4J). The best-rooted plants were transferred to soil and grown under greenhouse conditions (Fig. 4K). No morphological differences were observed among the regenerants, or between them and seed-initiated plants. All plants were found to be fertile and set viable seeds which germinated and produced morphologically normal plants.

Table 3.

Plant regeneration from 11-month-old embryogenic callus cultures^a

Cultivar	Cytokinin used for regenerationb	Calli forming green plantscd (%)	No of plants per embryogenic callus	Calli forming albinotic plants (%)
Golden Promise	0.1 mg l−1 BAP	55.5	2.6	0
	0.1 mg l−1 TDZ	60.0	3.0	0
Lomerit	0.1 mg l−1 BAP	60.0	1.4	0.03
	0.1 mg l−1 TDZ	67.8	2.2	0.04
Masto	0.1 mg l−1 BAP	54.8	2.0	0
	0.1 mg l−1 TDZ	59.0	2.8	0

a

Embryogenic callus was maintained on basal medium with 3% maltose+6 mg l⁻¹ 2,4-D+0.01 mg l⁻¹ BAP by biweekly subcultures.

b

Regeneration medium was basal medium with 3% maltose+1 mg l⁻¹ 2,4-D+above cytokinin.

c

Thirty pieces of embryogenic callus were used.

d

Calculated after two subcultures every 3 weeks on regeneration medium.

Somatic embryogenesis and plant regeneration from commercial cultivars

The protocol described above was established for cv. Golden Promise because of the high tissue culture ability of this genotype. After finding the appropriate hormone combination and culture conditions for high somatic embryogenesis, the protocol was applied to other barley genotypes having a high commercial value in Europe.

Concerning the induction of primary and embryogenic callus as well as plant regeneration, all cultivars showed a positive response (Table 4). After soaking the seeds in water containing 6 mg l⁻¹ 2,4-D and culturing on 6M6D, MEs of cv. Golden Promise showed more germination with less callus after 1 week of incubation, while least germination along with more callus formation was observed in cv. Borwina. Complete removal of the elongated main shoots and roots from the callus-responding MEs resulted in a high frequency of primary callus of more than 50% in all genotypes excluding cv. Salome. With respect to embryogenic callus formation, genotype specific differences were observed (Table 4). The cv. Masto gave the highest percentage followed by cvs Golden Promise, Duet, Borwina, and Franziska yielding similar frequencies, whereas a significantly lower percentage of callus was recorded for cvs Salome, Lomerit, and Merlot. Besides that, genotype-specific differences also appeared for the number of plants regenerated per embryogenic callus where cv. Masto again gave the best results. Interestingly, cv. Salome having the lowest frequency of embryogenic callus responded with a relatively high number of plants compared with some cultivars yielding higher embryogenic frequencies than cv. Salome. Occasionally, the regeneration of albino plants was observed, however, at very low ratios: 0 to 0.1 per embryogenic callus. Regenerated plants transferred to soil showed a normal phenotype and set seeds.

Table 4.

Frequency of primary and embryogenic callus induction as well as plant regeneration from mature embryos in different cultivars of barley^a

Cultivar	Mature embryos forming			Average number of plants per embryogenic callusde
	Primary callus (%)b	Embryogenic callus (%)ce	Green	Albino
Golden Promise	72.5±11	46.7±6.1 a	5.8±1.0 ae	0.10±0.03
Salome	41.3±7.5	22.3±4.1 bc	4.4±0.6 ac	0.10±0.09
Borwina	66.5±9.8	42.5±1.4 ad	1.5±0.1 b	0.04±0.05
Duet	72.9±6.9	45.1±5.6 ae	3.9±0.4 ad	0.03±0.03
Franziska	65.9±6.0	41.3±7.3 acf	1.4±0.3 b	0
Lomerit	50.5±10	25.9±12.3 bdef	2.3±0.5 bcd	0.08±0.13
Masto	74.6±4.9	55.1±2.1 a	7.5±1.1 e	0.04±0.03
Merlot	53.9±9.7	25.8±7.2 bdef	2.1±0.4 bd	0

a

Three independent experiments were carried out for each cultivar, for Golden Promise four repetitions were done, average data (±SE).

b

Frequency of primary callus induction was evaluated by counting the MEs that formed callus out of the total number of MEs cultured ×100.

c

Frequency of embryogenic callus formation was assessed by the percentage of MEs forming embryogenic callus and the total MEs explanted.

d

Plant regeneration ability was estimated as the total number of shoots/plantlets developed per embryogenic callus.

e

Values followed by the same letters are not significantly different at P=0.05 according to the ANOVA test.

Regeneration potential in long-term maintained cultures

Highly embryogenic callus from all the cultivars used could be maintained on 3M3DB or 3M6DB by biweekly subcultures. The regeneration ability of 11-month-old cultures selected randomly from cvs Golden Promise, Masto, and Lomerit was analysed, yielding a response in more than 50%. Well-developed and rooted plantlets could be regenerated upon two subcultures of embryogenic callus pieces (each 100–150 mg) in regeneration medium containing either 0.1 mg l⁻¹ BAP or TDZ (Table 3).

Discussion

A system for highly efficient and long-term somatic embryogenesis together with plant regeneration was developed for barley using mature dry seeds as the starting material. During the last two decades, MEs of cereals have been evaluated repeatedly as explants for embryogenic cultures, but with limited success, therefore immature embryos were preferred as explants for tissue culture because of their high regeneration potential. For barley, MEs were first described for callus induction by Bayliss and Dunn (1979) followed by a few reports demonstrating the ability to undergo somatic embryogenesis and subsequent plant regeneration, in principle.

The major obstacle to callus formation of MEs is germination, which is their normal and primary reaction. A combination of different mechanical and chemical approaches markedly reduced germination and favoured primary callus induction as a source for embryogenic callus formation in the present work. Injury or the complete removal of the elongated main shoots within a week of culture initiation redirected the development to callusing. Similar observations were reported by Bai and Qu (2001) who found a 6-fold increase in callus induction frequency by simply slicing the mature seeds of turf-type tall fescue. For wheat, a segmentation of MEs into small pieces was shown to improve somatic embryogenesis and to prevent embryo germination (Delporte et al., 2001). Furthermore, raising the amount of maltose in the induction medium caused a drastic reduction of embryo axis elongation with respect to the length of the main shoot in the present study. Similar observations concerning a reduction of germination of immature as well as mature embryos, improvement of embryogenic callus formation, and maintenance of long-term embryogenic capacity due to higher levels of sucrose, sorbitol, and maltose as well as salts have already been demonstrated (Galiba and Yamada, 1988; Ryschka et al., 1991; Barro et al., 1999). In addition, it was found that an early and close contact of the embryonic axis to 2,4-D was essential to reduce its elongation and root development. This was clearly shown by culturing the MEs with embryonic axis-side down and by isolation of MEs from the seeds preincubated in water containing 2,4-D. Moreover, the germination and callusing response of MEs turned out to be affected by raising the concentration of 2,4-D, reaching an optimum at 6 mg l⁻¹ combined with an increase in maltose that was most effective at 6% (w/v). Thus, these observations showed that a very fine balance between the growth regulator and the maltose concentration in the induction medium promotes callus induction. Besides that, the inclusion of BAP was also found to be highly beneficial for improved callus quality and regeneration ability for MEs, in accordance with earlier reports for immature embryos of barley (Cho et al., 1998; Dahleen and Bregitzer, 2002) and was further demonstrated in other cereal and grass species (Chaudhury and Qu, 2000; Huang and Wei, 2004). Unlike immature embryos, where the scutellar tissue was clearly identified as a source for embryogenesis (Magnusson and Bornman, 1985; Ryschka et al., 1991), the scutella of MEs did not show any response. Similar observations have already been made for barley by Oka et al. (1995) who reported callus formation and organogenesis with leaf primordia from the leaf epidermal tissue or coleoptile base of mature embryos as revealed histologically. However, in the present study, a much broader spectrum of growth regulators was analysed yielding the same findings concerning non-responding scutellar tissue, while plating the excised embryonic axes showed a better response than using whole MEs. Highly efficient somatic embryogenesis and plant regeneration arising from embryonic axes as shown in this study has not been reported previously for barley and other cereals. The question why the scutella of MEs did not give any callusing response remains to be analysed. In contrast to cereals, for herbaceous and woody legumes that are also known to be recalcitrant in tissue culture, the use of the embryonic axis as an explant for in vitro regeneration via organogenesis or embryogenesis was demonstrated several times (Zambre et al., 1998; Lacroix et al., 2003). Moreover, the suitability of the embryonic axis explants for transformation was shown (Krishnamurthy et al., 2000).

Recently, several publications have concentrated on the improvement of in vitro regeneration in cereals and grasses using MEs. For barley, Akula et al. (1999) reported a high cultivar dependency and showed that auxins inducing callus were cultivar specific, therefore media formulations had to be adapted to each genotype. Although this study's results revealed genotypic differences, it was successful in embryogenesis together with plant regeneration in seven additional cultivars with overall rates ranging from 22–55% after establishing the protocol for the model cv. Golden Promise. In contrast to this, Ganeshan et al. (2003) reported a very low induction frequency of embryogenic callus in MEs of barley. Comparable amounts of embryogenic callus formation, together with plant regeneration with less genotype dependency as described recently, were also demonstrated for MEs of maize (Huang and Wei, 2004) and mature seeds of japonica rice (Lee et al., 2002), whereas significant differences were found in wheat (Zale et al., 2004).

The suitability of barley MEs as explants for highly efficient plant regeneration through multiple shoot differentiation using Picloram and TDZ has previously been reported by Sharma et al. (2004). It is demonstrated here that the choice of growth regulators can direct the developmental process via somatic embryogenesis or organogenesis. In general, somatic embryogenesis is preferred for cereal biotechnology since plants derived from single cells avoiding the risk of selecting chimeras (Vasil, 1994).

In this report, a reproducible and highly efficient protocol is described, with long-term maintenance of morphogenic capacity and less genotype dependency based on mature dry barley seeds. As far as is known, this is the first report on the establishment of embryogenic cultures originating from embryonic axes in cereals. This system could pave the way for genetically transforming elite genotypes of barley independently of the gene transfer method chosen.

Abbreviations: BAP, 6-benzylaminopurine; 2,4-D, 2,4-dichlorophenoxyacetic acid; Dicamba, 3,6-dichloro-2-methoxybenzoic acid; IBA, indole-3-butyric acid; ME, mature embryo; Picloram, 4-amino-3,5,6-trichloropicolinic acid; TDZ, thidiazuron.

This work was supported by a grant through the ‘Forschungsschwerpunkt Agrarbiotechnologie des Landes Niedersachsen’. We are grateful to Christian Böhme and Claudia Hasse for excellent technical assistance and to Dr Volker Ssymank for statistical analysis. We thank Syngenta Seeds for providing seeds of cv. Franziska; Lochow-Petkus for cv. Lomerit; Limagrain Nickerson for cv. Duet; Semundo Saatzucht for cv. Masto; and Nordsaat Saatzuchtgesellschaft for cv. Merlot.