Somatic embryogenesis and the effect of particle bombardment on banana Maçã regeneration

A plant regeneration method with cell suspension cultures of banana, and the effect of biobalistic on regeneration potential are described in this report. Somatic embryos of banana were obtained from indirect embryogenesis of male inflorescence of banana cultivar Maçã (AAB group). Part of the calluses formed (40%) showed embryogenic characteristics (nonfriable, compact and yellow color). The cell suspension, originated from embryogenic calluses, contained clusters of small tightly packed cells with dense cytoplasms, relatively large nuclei and very dense nucleoli. After four months of culture, somatic embryos started to regenerate. The maximum number of regenerated plants was observed between 45 and 60 days after embryo formation. In the first experiment, 401 plants were regenerated from approximately 10 mL of packed cells. In the second experiment, 399 plants were regenerated from a cell suspension six months older than that of the first experiment. Cell transformation using particle bombardment with three different plasmid constructions, containing the uid-A gene, resulted in a strong GUS expression five days after bombardment; however, plant regeneration from bombarded cells was much lower than nonbombarded ones.


Introduction
More than 58 million tons of bananas and 30 million tons of plantains are produced worldwide (FAO, 2001).India is the largest producer of bananas (11 million tons), and Uganda is the largest producer of plantains (9 million tons).Nearly half of the world banana production is cultivated in Asia, while almost 75% of the world's plantains are cultivated in Africa.The world average yield is 15 tons ha -1 for banana and about 6 tons ha -1 for plantain.Banana is a staple food for nearly 400 million people throughout the developing world (Sasson, 1997).The main impediment for banana production is associated with fungal and viral diseases.
The development of new varieties by sexual crossing in banana is extremely difficult, because only few diploid clones produce viable pollen to use in crosses with female-fertile triploids for the production of a few tetraploid seeds (Shepherd, 1987).The major part of germplasm from the commercially most important clones (group AAA, AAB and ABB) is excluded from sexual breeding, because it is triploid and entirely male and female sterile (Novak et al., 1989).
In fact, the application of classical methods of breeding, for both disease and pest resistance, has resulted in limited success for banana.The integration of genetic engineering into breeding programs may provide powerful tools to overcome this limitation, by introducing specific genetic changes that can be utilized in banana improvement within a short period of time (Khalil et al., 2002).However, these applications require reliable plant callus or cell suspension regeneration protocols for banana.
Immature male inflorescences have been used to initiate cultures of several banana and plantain cultivars (Escalant et al., 1994;Sagi et al., 1995;Côte et al., 1996;Navarro et al., 1997).Escalant et al. (1994) suggested that somatic embryos are probably of unicellular origin.This method has the potential to produce nonchimeric plants, making it an excellent option for nonconventional breeding programs.Somatic embryogenesis techniques in the genus Musa are aimed at developing new, highperformance micropropagation techniques and a cell regeneration system useful for genetic improvement (Côte et al., 1996).In order to help breeding projects, the cell culture could be associated to mutation induction, transformation processes (biobalistic or Agrobacterium systems) or selection of material from somaclonal variation.Unlike meristem cultures, a large number of cells can be treated using cell suspension as explant.Published protocols for regeneration of banana plants from cell suspension have mainly focused on exporttype "Cavendish" (genome AAA).Therefore, the development of an efficient callus regeneration method for important Brazilian economic banana cultivars, such as cv.Maçã (Silk type, AAB group), could offer an alternative for crop improvement.
This research aimed at developing a plant regeneration method from callus, derived from cell suspension cultures from male inflorescences of banana cv.Maçã, and to study the effect of transformation, using microparticle bombardment, in the embryogenic callus regeneration potential.
The regenerated plants were transferred to the micropropagation medium (MM), which consisted of MS salts supplemented with BAP (5.0 mg L -1 ), sucrose (3 g L -1 ), Phytagel (2 g L -1 ).Part of the rooted material was transferred to a substrate and kept in an acclimatized glass house.Regenerated plants were maintained in a greenhouse to identify somaclonal variants.The entire regeneration process was performed twice, in order to evaluate the effect of aging on embryogenic viability of the cell suspension; the only difference was that cell suspension of the second experiment was six months older than that used in the first experiment.
For histological analysis, the embryogenic cell suspension was fixed in glutaraldehyde solution 2% (v/v) with paraformaldehyde 0.5% in a sodium cacodilato buffer (0.1 M, pH 7.2).The cell suspension was mixed with agar solution 4% and dropped in Petri dishes.A packed of cell suspension was incubated in ice for 5 minutes, fixed with osmium tetroxide solution 1% in sodium cacodilato buffer, and incubated overnight (uranila acetate 2.5%).After incubation, the blocked material was dehydrated in acetone and embedded in Epon 812 resin overnight.Sections of 0.4 mm were stained with 1% toluidine blue in 1% borate, and examined in light microscope.
After the establishment of somatic embryogenesis and plant regeneration protocols, some embryogenic calluses from cell suspension of banana cv.Maçã were bombarded with plasmid constructs, containing the gene for kanamycin resistance and the gus-A gene for β-glucoronidase.Ten Petri dishes were used containing 0.5 mL of packed cell suspension for each plasmid construction.The following constructs were used: pBI426 (70S promoter, uid-A/neo), pFF19 (70S promoter, uid-A) and pCAMBIA1303 (35S promoter, uid-A, Hygromycin).The DNA of each plasmid construction was isolated from the Escherichia coli DH5α strain by alkaline lysis, and purified by polyethylene glycol precipitation (Concert kit; Invitrogen).Tungsten particles were prepared using 5 µL of DNA (1 ng DNA µL -1 ), 20 µL of spermidine and 50 µL of CaCl 2 .All solutions were kept on ice.Calluses were placed 9 cm from the point of particle discharge, done with 1,100 psi in helium-driven particle gun for plant transformation.
One week after bombardment, embryogenic cells, including nonbombarded controls, were subcultured on embryogenic medium containing 100 mg L -1 kanamycin.Calluses were kept in the dark on embryo induction media (ME3) and subculturing to fresh media every 30 days.After initial embryo development, calluses were transferred to regeneration media (ME4) under light for plantlet growth induction.

Results and Discussion
During the first two weeks of culture in medium ME1, isolated immature male flower bracts developed necrosis.Callus formation was observed under some bracts after two months of culture initiation.Approximately 80% of the explants originated callus tissues.As Arnold et al. ( 2002) suggested, it is usually easy to distinguish between embryogenic and nonembryogenic calluses on the base of their morphology and color.Using their definition, 40% of banana calluses (type I) showed embryogenic characteristics such as: nonfriable, compact and yellow color (Figure 1 A), but 60% of calluses (type II) had nonembryogenic characteristics (friable and white) (Figure 1 B).
Both callus types were excised from the mother tissue and transferred to ME2 liquid medium.After three months, cell suspensions were established from both types.The cell shape depended completely upon the callus type.Cell suspension from callus type I exhibited small clusters of cells, which showed dense cytoplasms with small vacuoles and starch reserves in the form of small starch grains (Figure 1 C).Cell suspension from type II callus showed individual elongated cells (no clusters), with vacuolated cytoplasms and few starch reserves (Figure 1 D).According to Dhed´a et al. (1991), elongated and highly vacuolated cells apparently originate from the nonmeristematic tissue, while oval cells, containing large starch granules, originated from the meristematic tissue.The embryogenic cell suspension, originated from callus type I, contained clusters of small tightly packed cells with dense cytoplasms, a relatively large nuclei, very dense nucleoli and tiny starch and protein grains (Figure 1 E).These characteristics are generally associated with morphogenetic competence and characterize embryogenic cell masses (Dhed´a et al., 1991;Alemanno et al., 1996).Cell aggregates similar to those found in cell suspension type I were described before in other species, such as corn (Emons & Kieft, 1992) and wheat (Tabaeizadeh & Campeau, 1992).These cell clusters were called "proembryogenic masses" (Georget et al., 2000) and indicate that cell suspension has embryogenic potential.The cell suspension has a thick consistency when the cell concentration is high.Kosky et al. (2002) observed a similar result.However, regeneration induction is necessary to confirm the embryogenic potential.
Plant regeneration was observed only with the cell suspension originated from callus type I.This result was expected from embryogenic cells with characteristics described previously.According to Arnold et al. (2002), the competence for embryogenic induction may be the result of the varying auxin sensitivity of these cells.These authors also suggested that the developmental response pattern of cultured tissue is epigenetically determined and influenced by the stage of plant development and the nature of the explant.The banana cell suspension observed was characterized by a slow development of cell clusters, making the plant regeneration time-consuming.According to Georget et al. (2000), this is a common feature of many banana cell culture protocols.These authors suggested that signals, as cellular density, could participate in the development type of aggregates.
Embryogenic callus development was observed after one month of cell culture in semi-solid medium (Figure 2 A).All embryos emerged from callus tissue, and no secondary somatic embryo formation was observed.Cronauer-Mitra & Krikorian (1988) described similar results.In the first step of plant regeneration, embryogenic globular compact structures were identified, after three months of culture, in ME2 medium (Figure 2 B).The development of these globular structures originated an apical invaginated area, corresponding to the cotyledonary stage of zygotic embryos.Differences in somatic stages of embryogenesis were observed on the callus surface.
During the regeneration time, the maximum number of regenerated plants was observed between 45 and 60 days of culture, in two experiments.In the first experiment (I), 401 plants were regenerated from approximately 10 mL of packed embryogenic cells.In the second experiment (II), using the same initial cell volume, 399 plants were regenerated from a cell suspension that was six months older than that of the first experiment (Table 1).In most crops, the embryogenic potential decreases with extended culture and is eventually lost (Arnold et al., 2002).The germination of embryos initiated with shoot development and leaf differenciation (Figure 2 C and D), followed by the elongation of the primary roots and the formation of adventitious roots (Figure 2 E).After embryo convertion, no embryo development was observed in embryogenic callus.
Some plants from both experiments showed slower development, when compared with plants of the same age regenerated from shoot tips of banana cv.Maçã.These growth differences were also observed, when the plants were transferred to the rooting medium (Figure 3 A).After acclimatization, these plants continued to show low height (Figure 3 B and C), probably as a result of somaclonal variation, often described in plants regenerated from cell culture.Arnold et al. (2002) suggested that mutation is the origin of this kind of off-type, because mutations could affect genes that play a more direct role in plant growth and development.Only mature embryos with a normal morphology, and which have accumulated enough storage material, and acquired desiccation tolerance at the end of maturation, could develop into normal plants.Somaclonal variation is a problem caused, in many cases, either by a prolonged callus phase and the use of 2,4-D (Arnold et al., 2002).
There is no information concerning the exact ontogenesis of these plants from callus.However, it may be suggested that they originated from somatic embryogenesis, because the sequential steps observed here are similar to those described for other monocot species (Conceição et al., 1998).
Transient expression was confirmed by GUS histochemical assay five days after bombardment.All constructs were GUS-positive.Similar levels of expression where observed, when the maize ubiquitin or the duplicated 35S CaMV promoter were used (Figure 4).Embryogenic structures were observed six months after bombardment, and plant regeneration was observed in all materials three months after transferring to light conditions.The number of plants obtained from bombarded material was inferior to nonbombarded embryogenic cells.The best result was observed with pBI426, in which 29 plants were regenerated, while only two plants were regenerated from callus bombarded with pCAMBIA1303.The reduction of the regenerating potential has been described before in bombarded banana cell suspension (Becker et al., 2000).
These results can be an effect of the antibiotic selection and bombardment treatment.Becker et al. (2000) reported that selection of bombarded embryogenic cell suspension of Bluggoe (Musa ssp.ABB group), with geneticin, allowed transgenic embryo transformation but interfered with embryo germination.This inhibitory effect has also been previously reported in other crops (Yao et al., 1995;Bretagnesagnard & Chupeau, 1996).Shock waves and sound waves, promoted by the biobalistic equipment, can cause damage to the explant target at the cellular level, as membranes disrupt, thus interfering with the regenerative process (Russel et al., 1992).

Figure 3 .
Figure 3. Aspects of banana cv.Maçã plants obtained from somatic embryogenesis.A: rooted plants regenerated from cell suspension with one month (yellow bar = 1.5 cm); B: acclimatized plants regenerated from cell suspension with size differences; C: acclimatized plants regenerated from cell suspension and from shoot tips with the same age (yellow bar = 10 cm).

Figure 4 .
Figure 4. Aspect of cell suspension transformation of banana cv.Maçã.A: nontransformed cell suspension; B and C: positive GUS test in embryogenic cell cluster; D: somatic embryogenesis of callus formed from bombarded cell suspension.

Table 1 .
Regenerated plants obtained from cell suspension with different culture period.