PlantCell Reports
Plant Cell Reports (1993) 13:41-44
9 Springer-Verlag 1993
Plant regeneration from cultured protoplasts of the cooking banana cv. Bluggoe (Musa spp., ABB group) R. Megia t, R. Ha'icour 1, S. Tizroutine 1, V. Bui Trang 1, L. Rossignol 1, D. Sihachakr 1, and J. Schwendiman 1 Morphogtn~se Vtgttale Exptrimentale, B~t. 360, Universit6 Paris Sud, F-91405 Orsay Cedex, France 2 BIOTROP - I R F A / C I R A D , B.P. 5035, F-34032 Montpellier Cedex, France Received 5 April 1993/Revised version received 9 June 1993 - Communicated by A. M. Boudet
Summary. Suspensions of embryogenic cells of a triploid banana (Musa spp., cv. Bluggoe) were initiated from the uppermost part of meristematic buds, and used as protoplast source. After 20 weeks in culture, the suspension contained a mixture of globular structures or globules and embryogenic cell clusters, as well as single cells. Two types of protoplasts were obtained from embryogenic suspension culture: small (20-30 l.tm) and larger (30-50 gm) protoplasts with a dense cytoplasm and large starch grains respectively. The small protoplasts probably originated from embryogenic cell clusters, and also from pseudocambial cells of globules, while larger protoplasts were probably released from oval starchy cells and those of the globule peripheral area. In co-culture with a suitable feeder, consisting of suspensions of diploid banana cells, the protoplasts of triploid banana reformed the cell wall within 24 h and underwent sustained divisions leading to the formation of small clusters of 2-3 cells within 7 days. The latter developed directly into embryos without passing through an apparent callus phase. 10% of such embryos gave rise to plantlets when subcultured in 2.2 pM 6benzylaminopurine and 2 ~tM 4 amino-3,5,6trichloropicolinic acid for 1 week, before transfer to MS medium containing 10 pM 6-benzylaminopurine. The rest of the embryos underwent intensive direct secondary embryogenesis which could lead to the formation of plantlets with a frequency of up to 50% upon further transfer to hormone-free medium. Key
words: Musa spp. - Cooking banana - Cell suspension - Somatic embryo - Protoplast
Abbreviations: BAP: 6-benzylaminopurine; MS: Murashige and Skoog (1962) medium; 2,4-D: dichlorophenoxyacetic acid; UV: ultraviolet light; FDA: fluorescein diacetate; MES: 2-(N-morpholino)ethanesulfonic acid; Picloram: 4 amino-3,5,6-trichloropicolinic acid.
from the family Gramineae (Pennisetum americanum, Vasil and Vasil 1980; Oryza sativa, Lee et al. 1989; Zea mays, Shillito et al. 1989; Hordeum vulgate, Funatsuki et al. 1992). Nevertheless, very few successful protoplast cultures have been reported for banana in particular (Megia et at. 1992). The high sterility of cultivated triploid bananas has resulted in very limited improvements of these plants by classical breeding methods. Therefore, tissue culture techniques, including cell fusion and transformation, are being employed in conjunction with classical breeding methods for the improvement of bananas. The plants appeared to be recalcitrant, and the culture of their protoplasts has proved difficult. In fact, while protoplasts can easily be isolated from various organs or cell suspensions of banana (Bakry 1984; Chen and Ku 1985; Da Silva Conceicao 1989), sustained divisions leading to callus formation were only obtained when embryogenic cell suspension-derived protoplasts were co-cultured at high density with very active feeder ceils (Megia et al. 1992). In banana, such embryogenic cell suspensions can be established from immature zygotic embryos (Escalant and Teisson 1989; Marroquin et al. 1993), meristematic corm tissues (Novak et al. 1989), meristematic shoot tips (Dhed'a et at. 1991), and immature male flowers (Escalant et at. 1992). The present study was, therefore, undertaken with the aim of developing a reproducible protocol for plant regeneration from cultured protoplasts of banana through somatic embryogenesis. Moreover, this study describes cytologically the origin of embryogenic cells used as protoplast source, and the early stages of the development of cultured protoplasts, as well as their differentiation into somatic embryos.
Material and methods Introduction In the past decade, significant progress in protoplast technology has resulted in regenerating whole plants from isolated protoplasts of monocotyledonous species, mainly Correspondence to: L Rossignol
Plant materials. Experiments were carried out with the very highly proliferating culture of shoot tips of the triploid cooking banana (ABB group), cv. "Bluggoe", kindly provided by the Katholieke Universiteit Leuven. Embryogemc cell suspension, used as protoplast source, was initiated by using the uppermost part (scalps) of meristematic bndsl which had been excised from a 3-week-old culture. About 60 scalps (4 mm in diameter each) were suspended in 20 ml of a liquid 1/2 MS basal medium
42 supplemented with vitamins (Morel and Wetmore 1951), 5 ].tM 2,4-D and ll.tM zeatin. The cultures were kept in 100 ml Erlenmeyer flasks which were placed on a gyratory shaker at 70 rpm, and illuminated with a photoperiod of 16h/day at 30 [.tM/m2/s and 27~ After the old-culture medium was replaced twice with a fresh medium at 8-week-intervals, the remaining explants were removed by filtration through a 0.5 mm nylon sieve. The resulting suspension contained isolated cells and globular structures. Subsequent subculture was done at 2-3-week intervals by diluting the suspension cultures twice in fresh medium. The embryogenic cell dusters appeared after 5 months of culture. The cell viability was regularly determined by staining with 10 I.tM FDA (Widhohm 1972). The cell feeders for protoplast culture were diploid cell suspensions of Musa acuminata ssp. burmannica "Long Tavoy" (AA group), kindly provided by Dr. J.V.Escalant (CIRAD, Montpellier, France).
Protoplast isolation and culture of protoplasts and feeder cells. The protocol for protoplast isolation and culture, as well as the establishment of the feeder cell cultures as described in Megia et al. (1992) were used. The enzyme mixture was composed of 1.5 % (w/v) cellulase RS (Yakult Honsha Co, Japan), 0.15% (w/v) Pectolyase Y23 (Sheishin Co, Japan), 0.2 % (w/v) hemicellulase (Sigma Co, USA), 3% (w/v) KC1, 0.5% (w/v) CaC12, and the pH adjusted to 5.6. The cell wall degradation was visualized through an UV inverted microscope after calcofluor white (10.4 [.tM) was added to the culture medium (Galbraith 1981). Protoplast viability was determined by staining with FDA (Widhohm 1972). Protoplasts were suspended in culture medium at a high density of 5 x 105 protoplasts/ml. The culture medium was N6 salts ( C h u e t al. 1975) supplemented with vitamins, organic acids and sugar alcohol (Kao and Michayluk 1975), 0.35 M glucose and 0.12 M sucrose as osmoticum, 1.9 mM KH2PO 4 and 0.5 mM MES.
Plant regeneration. Embryogenic clusters and embryos at various stages which derived from the cultured protoplasts were transferred to the regeneration medium composed of MS basal medium but with macroelements used at haft strength. The medium was supplemented with vitamins (Morel and Wetmore 1951), 87.7 mM sucrose, 10 pM BAP used alone or combined with Picloram, and solidified with 2 g/l gelrite. In order to stimulate plantlet development, germinating embryos were then transferred to the same medium but lacking growth regulators and containing 58.5 mM sucrose.
Morphological and histological analysis. For histological examination, samples of cells were taken from the suspensions prior to enzyme digestion, and those of cultured protoplasts after 1, 2 and 7 weeks of culture. Samples were embedded in 2% (w/v) agar, and fixed in a 0.2 M phosphate buffer (pH 7.2) containing 2.5% (w/v) formaldehyde, 1% (v/v) glutaraldehyde and 1% (w/v) cafeine for 24 h at 4 ~ C. The fixed samples were dehydrated in ethanol, and embedded in historesin. Sections of 2 btm thick were stained as described by Megia et al. (1992).
Results and discussion
Initiation and maintenance of cell suspensions After one week of culture, the inoculated "scalp" explants lost their epidermis and morphological organization, as described by Dhed'a et al. (1991). They consisted of meristematic and non-meristematic tissues (Sannasgala 1989). After 4-5 weeks of culture, white yellowish globular structures which appeared at the surface of the "scalps", and 50-120 /am single cells with an apparent nucleus were released into the culture medium. Predominantly, elongated and highly vacuolated cells, originating from nonmeristematic tissues of the "scalps" were released early into the medium. After 20 weeks in culture, the suspension was heterogeneous and contained a mixture of globular slructures at different stages and cell clusters, as well as single cells. Particularly, 3 types of cells can be observed. The first type was identified as elongated and highly vacuolated cells which had been released early into the medium, and the second type as oval-shaped cells (75-100/am) with a dense cytoplasm and large starch grains. The third cellular type consisted of spherical cells, small in size (20-30 /am) and having embryogenic characteristics (a dense cytoplasm with a great
amount of reserves, particularly small grains of starch and proteins, a large nucleus with a large unique nucleolus and a rather thick cell wall). These cells were rarely found as single cells, but often formed very small clusters of 3-15 cells and 50-250/am in diameter. Frequent subcultures of the suspensions into new medium stimulated the growth rate of globular structures and embryogenic cell clusters, leading to cell suspension which became both granular and milky. In the meantime, the number of elongated and highly vacuolated cells was decreased. Cytological examination showed that globular structures or globules which derived from perivascular regions of scalps (Sannasgala 1989) proliferated by budding. One or several buds appeared at the surface of primary globules, then, underwent rapid growth, and their fragmentation occurred following the preexisting cleavage lines. A well-formed globule (Fig. 1A) was composed of 3 different zones: the center, the pseudocambial and peripheral areas (Fig. 1A, B). The center formed by differentiated cells was surrounded by a pseudocambial area which ensured the globule growth. At the surface of this area, cells divided actively and accumulated starch progressively, leading to the formation of the globule amylaceous area which first released oval-shaped cells with large starch grains. Then, the peripheral ceils of this zone, which displayed a large amount of amylaceous reserves but without any apparent nuclear activity, underwent a progressive hydrolysis of their reserves, and finally differentiated into cells with embryogenic characteristics (Fig. 1B). The large globules can be purified by filtration of the suspensions through a 1 mm nylon mesh, but the separation of different types of cells was unsuccessful. The suspensions were, however, gradually enriched with embryogenic cell clusters after successive subcultures in fresh medium. Moreover, plant regeneration can be obtained from embryogenic cells with a frequency of 15%. Such embryogenic cell suspensions were then used as protoplast source for further successful plant regeneration as described below. Protoplast isolation and culture The enzyme composition previously used for diploid cell suspensions of banana (Megia et al. 1992) was ineffective in isolating protoplasts from triploid banana cells, probably due to the high concentration of enzymes used, causing lysis of protoplasts. Therefore, the reduction of cellulase and pectinase concentrations, combined with the addition of hemicellulase resulted in a good yield of protoplasts. The isolation of protoplasts was optimal, when embryogenic cell suspensions of 7-10 days after subculture were used as source of protoplasts, yielding 6-15 x 105 protoplasts/g fr. wt. Both a truly spherical shape of the released protoplasts, and the absence of fluorescence after staining with calcofluor white confirmed the complete cell wall degradation. Two types of protoplasts were obtained (Fig. 1C): small (20-30 /am) and larger (30-50 /am) protoplasts with a dense cytoplasm and large starch grains respectively. The small protoplasts probably originated from embryogenic cell clusters, and also from pseudocambial cells of globules after enzyme digestion, while larger protoplasts were probably released from oval starchy cells and those of the globule peripheral area, rich in amylaceous reserves. Test of
43
Fig.1. A) A globule with 3 different zones: the center (C), the pseudocambial (Ps) and peripheral (Pr) areas; B) Cytological analysis showing embryogenie cells (F_,c)at the surface, and peripheral cells (Pc) beneath; C) Two types of protoplasts: small (Sp) and larger (Lp) protoplasts with a dense cytoplasm and large starch grains respectively; D) A protoplast-derived embryogenic cell cluster;, E) A protoplast-derived globular embryo; F) Somatic embryos derived from cultured protoplasts of banana; G) Section of an immature somatic embryo with a root axis (Ra), a haustorium (Ha) and an epidermis (Ep), but lacking the shoot axis; H) Secondary embryos; I) A planflet regenerated from cultured protoplasts via somatic embryogenesis; J) Plants regenerated from banana protoplasts and transplanted in soil.
44 viability by FDA staining revealed that 90% of the protoplasts were viable as shown by yellow-green fluorescence. In co-culture with a suitable feeder, consisting of suspensions of very active cells of diploid banana, the protoplasts of triploid banana reformed the cell wall within 24 h, and underwent sustained divisions.
Plant regenerationfrom culturedprotoplasts After 3 days of co-culture with the feeder cells, first divisions occurred, leading to the formation of small clusters of 2-3 cells within 7 days. It has been shown earlier that the presence of active nurse cells was required to induce sustained divisions of protoplasts, which did not divide otherwise or rarely did (Megia et al. 1992). In this study, the plating efficiency ranged from 0.35 to 0.65% on day 10. The cell clusters of all sizes can be observed by the naked eye on day 14 (Fig. 1D). They underwent sustained cell divisions and gave rise to globular embryos of various sizes, and the older ones had a well-formed epidermis after day 50 (Fig. 1E). The somatic embryos developed a suspensor and a haustorium (Fig. 1F, G). No apparent callus phase was observed throughout the development of banana protoplastderived cell clusters into globular embryos. Further development into plants was obtained when protoplastderived embryos at globular and cotyledonous stages were incubated on regeneration medium. Only 3% of them gave rise to shoots and plantlets after 5 weeks of culture in 10 ~tM BAP (Fig. 1I). The frequency of plant regeneration was improved up to 10% by subculturing embryos in 2.2 lxlVI BAP and 2 ktM Picloram for one week, before transfer to MS medium containing 10 ~tM BAP. It indicated that most of the globular structures did not bear any shoot axis, since they were not yet bipolar, and a period of maturation was necessary to obtain a bipolar structure able to germinate. The rest of the primary embryos, cultured on medium with 10 ~tM BAP, underwent intensive direct secondary embryogenesis (Fig. 1H). This occurred from the upper layer of primary embryos, leading to the formation of multiple large somatic embryos and small single globular embryos. The multiple large somatic embryos continued to grow, and developed into shoots and plantlets with a frequency of up to 50% within 2 months, while only 7% of small embryos gave rise to plantlets on hormone-free medium. Secondary globular embryos were morphologically identical to primary embryos. Under appropriate conditions, they underwent either a further cycle of embryogenesis or plant development. Sufficiently developed shoots and plantlets were rooted on hormone-free MS medium before transfer to the greenhouse (Fig. 1J). As far as plant regeneration was concerned, the ontogenetic state of the explant source, and particularly embryogenic suspension cultures have proved determinant for successful protoplast cultures in monocotyledonous species (Vasil 1988). A reproducible protocol for culture of banana protoplasts has previously been developed (Megia et al. 1992). Plant regeneration has not, however, been achieved, probably due to the loss of totipotency of the old culture of diploid cell suspensions which had been used as protoplast source. In this study, the possibility that plantlet formation might have regenerated from the diploid cell suspensions used as feeder cells, was excluded because of
their inability to regenerate shoots. On the contrary, the use of freshly initiated embryogenic suspensions as protoplast source, displaying morphogenetic potential, has led to successful plant regeneration. Fast-growing and predominantly embryogenic suspensions cultures were highly desirable, as the system provided competent protoplasts (Vasil and Vasil 1987; Kyozuka et al. 1987; Funatsuki et al. 1992). However, it took a long time (at least 6 months) to establish such embryogenic suspensions of triploid banana, while at the same time suspensions tended to lose cell totipotency as they aged. In order to maintain morphogenic potential, it has been suggested either to routinely reinitiate cultures at intervals (Vasil 1988), or to use the technique of cryopreservation (Panis et al. 1990). In this study, the establishment of reproducible protoplast to plant system has successfully been achieved. However, the application of new techniques, including protoplast fusion and transformation, requires cultures of protoplasts from other genotypes of banana. Herein it has been shown that plants have directly been regenerated from cultured protoplasts via somatic embryogenesis, without passing through an apparent callus phase. Consequently, the morphogenetic events resulting in direct initiation of organized meristematic tissues may favour genetic stability (D'Amato 1975). Further morphological and cytological analyses, as well as field evaluation are needed for verifying the conformity of regenerated plants. Acknowledgements: This study was supported by a grant from EC/STD3. The authors wish to thank Dr. J.V. Escalant and the Laboratory Tropische Plantenteelt Leuven (K.U.L., Belgium) for kindly providing plant material, M. D. Froger for the photography.
References Bakry F (1984) Fruits 39:449-452 Chen WH, ZC Ku ZC (1985) J Agric Assoc China 129:56-57 Chu CC, Wang CC, Sun C.S. (1975) Sci Sinica 5:659-668 D'Amato F (1975) In: Frankel OH, Hawkes JG (eds) Crop Genetic Resources for Today and Tomorrow, Cambridge Unv. Press, pp 333348. Da Silva Conceicao A (1989) DEA G~n et S~lec Animale et Vegetale, Univ. Paris Sud, 14p Dhed'a D, Dmnortier F, Panis B, Vuylsteke D, De Langhe E (1991) Fruits 48:125-135 Escalant JV, Teisson C (1989) Plant Cell Rep 7:665-668 Escalant JV, Babeau J, Chatelet C, Teisson C (1992) In: Wills B (ed) Proc Inter Symp Genet. Improvement of bananas for resistance to diseases and pests, Montpellier - France, 7-9 Sept 1992 Funatsuki H, Ltrz H, Lazzeri PA (1992) Plant Sci 85:179-187 Galbraith DW (1981) Physiol Plant 53:111 - 116 Kao KN, Michaylnk MR (1975) Planta 126:105-110 Kyozuka J, Hayashi Y, Shimamoto K (1987) Mol Gen Genet 206: 408413 Lee L, Behroll RE, Grimes ND, Hodges (1989) Planta 178:325-333 Marroquin CG, Paduscheck C, Teisson C (1993) In Vitro Cell Dev Biol 29:43-46 Megia R, HaYcourR, Rossignol L, Sihachakr D (1992) Plant Sci 85:91-98 Morel G, Wetmore RH (1951) Am J Bot 38:141-143 Murashige T, Skoog F (1962) Physiol Plant 15:473-497 Novak FJ, Afza R, Van Duren M, Perea-DaUos M, Conger BV, Tang Xiao Lang (1989) Biotechnology 7:154-159 Panis BJ, Withers LA, De Langhe E (1990) Cryo Letters 11:337-350 Sannasgala AK (1989) PhD Thesis, Katholieke Univ Leuven, Belgium, 172 p Shillito RD, Carswell GK, Jonhson CM, Di Maio JJ, Harms CT (1989) Biotechnology 7:581-587 Vasil IK (1988) Biotechnology 6:397-402 Vasil V, Vasil IK (1980) Theor Appl Genet 56:97-99 Vasil V, Vasil IK (1987) Theor Appl Genet 73:793-798 Widhokn JM (1972) Stain Techno147:189-194
Plant Cell Reports (1993) 13:41-44
9 Springer-Verlag 1993
Plant regeneration from cultured protoplasts of the cooking banana cv. Bluggoe (Musa spp., ABB group) R. Megia t, R. Ha'icour 1, S. Tizroutine 1, V. Bui Trang 1, L. Rossignol 1, D. Sihachakr 1, and J. Schwendiman 1 Morphogtn~se Vtgttale Exptrimentale, B~t. 360, Universit6 Paris Sud, F-91405 Orsay Cedex, France 2 BIOTROP - I R F A / C I R A D , B.P. 5035, F-34032 Montpellier Cedex, France Received 5 April 1993/Revised version received 9 June 1993 - Communicated by A. M. Boudet
Summary. Suspensions of embryogenic cells of a triploid banana (Musa spp., cv. Bluggoe) were initiated from the uppermost part of meristematic buds, and used as protoplast source. After 20 weeks in culture, the suspension contained a mixture of globular structures or globules and embryogenic cell clusters, as well as single cells. Two types of protoplasts were obtained from embryogenic suspension culture: small (20-30 l.tm) and larger (30-50 gm) protoplasts with a dense cytoplasm and large starch grains respectively. The small protoplasts probably originated from embryogenic cell clusters, and also from pseudocambial cells of globules, while larger protoplasts were probably released from oval starchy cells and those of the globule peripheral area. In co-culture with a suitable feeder, consisting of suspensions of diploid banana cells, the protoplasts of triploid banana reformed the cell wall within 24 h and underwent sustained divisions leading to the formation of small clusters of 2-3 cells within 7 days. The latter developed directly into embryos without passing through an apparent callus phase. 10% of such embryos gave rise to plantlets when subcultured in 2.2 pM 6benzylaminopurine and 2 ~tM 4 amino-3,5,6trichloropicolinic acid for 1 week, before transfer to MS medium containing 10 pM 6-benzylaminopurine. The rest of the embryos underwent intensive direct secondary embryogenesis which could lead to the formation of plantlets with a frequency of up to 50% upon further transfer to hormone-free medium. Key
words: Musa spp. - Cooking banana - Cell suspension - Somatic embryo - Protoplast
Abbreviations: BAP: 6-benzylaminopurine; MS: Murashige and Skoog (1962) medium; 2,4-D: dichlorophenoxyacetic acid; UV: ultraviolet light; FDA: fluorescein diacetate; MES: 2-(N-morpholino)ethanesulfonic acid; Picloram: 4 amino-3,5,6-trichloropicolinic acid.
from the family Gramineae (Pennisetum americanum, Vasil and Vasil 1980; Oryza sativa, Lee et al. 1989; Zea mays, Shillito et al. 1989; Hordeum vulgate, Funatsuki et al. 1992). Nevertheless, very few successful protoplast cultures have been reported for banana in particular (Megia et at. 1992). The high sterility of cultivated triploid bananas has resulted in very limited improvements of these plants by classical breeding methods. Therefore, tissue culture techniques, including cell fusion and transformation, are being employed in conjunction with classical breeding methods for the improvement of bananas. The plants appeared to be recalcitrant, and the culture of their protoplasts has proved difficult. In fact, while protoplasts can easily be isolated from various organs or cell suspensions of banana (Bakry 1984; Chen and Ku 1985; Da Silva Conceicao 1989), sustained divisions leading to callus formation were only obtained when embryogenic cell suspension-derived protoplasts were co-cultured at high density with very active feeder ceils (Megia et al. 1992). In banana, such embryogenic cell suspensions can be established from immature zygotic embryos (Escalant and Teisson 1989; Marroquin et al. 1993), meristematic corm tissues (Novak et al. 1989), meristematic shoot tips (Dhed'a et at. 1991), and immature male flowers (Escalant et at. 1992). The present study was, therefore, undertaken with the aim of developing a reproducible protocol for plant regeneration from cultured protoplasts of banana through somatic embryogenesis. Moreover, this study describes cytologically the origin of embryogenic cells used as protoplast source, and the early stages of the development of cultured protoplasts, as well as their differentiation into somatic embryos.
Material and methods Introduction In the past decade, significant progress in protoplast technology has resulted in regenerating whole plants from isolated protoplasts of monocotyledonous species, mainly Correspondence to: L Rossignol
Plant materials. Experiments were carried out with the very highly proliferating culture of shoot tips of the triploid cooking banana (ABB group), cv. "Bluggoe", kindly provided by the Katholieke Universiteit Leuven. Embryogemc cell suspension, used as protoplast source, was initiated by using the uppermost part (scalps) of meristematic bndsl which had been excised from a 3-week-old culture. About 60 scalps (4 mm in diameter each) were suspended in 20 ml of a liquid 1/2 MS basal medium
42 supplemented with vitamins (Morel and Wetmore 1951), 5 ].tM 2,4-D and ll.tM zeatin. The cultures were kept in 100 ml Erlenmeyer flasks which were placed on a gyratory shaker at 70 rpm, and illuminated with a photoperiod of 16h/day at 30 [.tM/m2/s and 27~ After the old-culture medium was replaced twice with a fresh medium at 8-week-intervals, the remaining explants were removed by filtration through a 0.5 mm nylon sieve. The resulting suspension contained isolated cells and globular structures. Subsequent subculture was done at 2-3-week intervals by diluting the suspension cultures twice in fresh medium. The embryogenic cell dusters appeared after 5 months of culture. The cell viability was regularly determined by staining with 10 I.tM FDA (Widhohm 1972). The cell feeders for protoplast culture were diploid cell suspensions of Musa acuminata ssp. burmannica "Long Tavoy" (AA group), kindly provided by Dr. J.V.Escalant (CIRAD, Montpellier, France).
Protoplast isolation and culture of protoplasts and feeder cells. The protocol for protoplast isolation and culture, as well as the establishment of the feeder cell cultures as described in Megia et al. (1992) were used. The enzyme mixture was composed of 1.5 % (w/v) cellulase RS (Yakult Honsha Co, Japan), 0.15% (w/v) Pectolyase Y23 (Sheishin Co, Japan), 0.2 % (w/v) hemicellulase (Sigma Co, USA), 3% (w/v) KC1, 0.5% (w/v) CaC12, and the pH adjusted to 5.6. The cell wall degradation was visualized through an UV inverted microscope after calcofluor white (10.4 [.tM) was added to the culture medium (Galbraith 1981). Protoplast viability was determined by staining with FDA (Widhohm 1972). Protoplasts were suspended in culture medium at a high density of 5 x 105 protoplasts/ml. The culture medium was N6 salts ( C h u e t al. 1975) supplemented with vitamins, organic acids and sugar alcohol (Kao and Michayluk 1975), 0.35 M glucose and 0.12 M sucrose as osmoticum, 1.9 mM KH2PO 4 and 0.5 mM MES.
Plant regeneration. Embryogenic clusters and embryos at various stages which derived from the cultured protoplasts were transferred to the regeneration medium composed of MS basal medium but with macroelements used at haft strength. The medium was supplemented with vitamins (Morel and Wetmore 1951), 87.7 mM sucrose, 10 pM BAP used alone or combined with Picloram, and solidified with 2 g/l gelrite. In order to stimulate plantlet development, germinating embryos were then transferred to the same medium but lacking growth regulators and containing 58.5 mM sucrose.
Morphological and histological analysis. For histological examination, samples of cells were taken from the suspensions prior to enzyme digestion, and those of cultured protoplasts after 1, 2 and 7 weeks of culture. Samples were embedded in 2% (w/v) agar, and fixed in a 0.2 M phosphate buffer (pH 7.2) containing 2.5% (w/v) formaldehyde, 1% (v/v) glutaraldehyde and 1% (w/v) cafeine for 24 h at 4 ~ C. The fixed samples were dehydrated in ethanol, and embedded in historesin. Sections of 2 btm thick were stained as described by Megia et al. (1992).
Results and discussion
Initiation and maintenance of cell suspensions After one week of culture, the inoculated "scalp" explants lost their epidermis and morphological organization, as described by Dhed'a et al. (1991). They consisted of meristematic and non-meristematic tissues (Sannasgala 1989). After 4-5 weeks of culture, white yellowish globular structures which appeared at the surface of the "scalps", and 50-120 /am single cells with an apparent nucleus were released into the culture medium. Predominantly, elongated and highly vacuolated cells, originating from nonmeristematic tissues of the "scalps" were released early into the medium. After 20 weeks in culture, the suspension was heterogeneous and contained a mixture of globular slructures at different stages and cell clusters, as well as single cells. Particularly, 3 types of cells can be observed. The first type was identified as elongated and highly vacuolated cells which had been released early into the medium, and the second type as oval-shaped cells (75-100/am) with a dense cytoplasm and large starch grains. The third cellular type consisted of spherical cells, small in size (20-30 /am) and having embryogenic characteristics (a dense cytoplasm with a great
amount of reserves, particularly small grains of starch and proteins, a large nucleus with a large unique nucleolus and a rather thick cell wall). These cells were rarely found as single cells, but often formed very small clusters of 3-15 cells and 50-250/am in diameter. Frequent subcultures of the suspensions into new medium stimulated the growth rate of globular structures and embryogenic cell clusters, leading to cell suspension which became both granular and milky. In the meantime, the number of elongated and highly vacuolated cells was decreased. Cytological examination showed that globular structures or globules which derived from perivascular regions of scalps (Sannasgala 1989) proliferated by budding. One or several buds appeared at the surface of primary globules, then, underwent rapid growth, and their fragmentation occurred following the preexisting cleavage lines. A well-formed globule (Fig. 1A) was composed of 3 different zones: the center, the pseudocambial and peripheral areas (Fig. 1A, B). The center formed by differentiated cells was surrounded by a pseudocambial area which ensured the globule growth. At the surface of this area, cells divided actively and accumulated starch progressively, leading to the formation of the globule amylaceous area which first released oval-shaped cells with large starch grains. Then, the peripheral ceils of this zone, which displayed a large amount of amylaceous reserves but without any apparent nuclear activity, underwent a progressive hydrolysis of their reserves, and finally differentiated into cells with embryogenic characteristics (Fig. 1B). The large globules can be purified by filtration of the suspensions through a 1 mm nylon mesh, but the separation of different types of cells was unsuccessful. The suspensions were, however, gradually enriched with embryogenic cell clusters after successive subcultures in fresh medium. Moreover, plant regeneration can be obtained from embryogenic cells with a frequency of 15%. Such embryogenic cell suspensions were then used as protoplast source for further successful plant regeneration as described below. Protoplast isolation and culture The enzyme composition previously used for diploid cell suspensions of banana (Megia et al. 1992) was ineffective in isolating protoplasts from triploid banana cells, probably due to the high concentration of enzymes used, causing lysis of protoplasts. Therefore, the reduction of cellulase and pectinase concentrations, combined with the addition of hemicellulase resulted in a good yield of protoplasts. The isolation of protoplasts was optimal, when embryogenic cell suspensions of 7-10 days after subculture were used as source of protoplasts, yielding 6-15 x 105 protoplasts/g fr. wt. Both a truly spherical shape of the released protoplasts, and the absence of fluorescence after staining with calcofluor white confirmed the complete cell wall degradation. Two types of protoplasts were obtained (Fig. 1C): small (20-30 /am) and larger (30-50 /am) protoplasts with a dense cytoplasm and large starch grains respectively. The small protoplasts probably originated from embryogenic cell clusters, and also from pseudocambial cells of globules after enzyme digestion, while larger protoplasts were probably released from oval starchy cells and those of the globule peripheral area, rich in amylaceous reserves. Test of
43
Fig.1. A) A globule with 3 different zones: the center (C), the pseudocambial (Ps) and peripheral (Pr) areas; B) Cytological analysis showing embryogenie cells (F_,c)at the surface, and peripheral cells (Pc) beneath; C) Two types of protoplasts: small (Sp) and larger (Lp) protoplasts with a dense cytoplasm and large starch grains respectively; D) A protoplast-derived embryogenic cell cluster;, E) A protoplast-derived globular embryo; F) Somatic embryos derived from cultured protoplasts of banana; G) Section of an immature somatic embryo with a root axis (Ra), a haustorium (Ha) and an epidermis (Ep), but lacking the shoot axis; H) Secondary embryos; I) A planflet regenerated from cultured protoplasts via somatic embryogenesis; J) Plants regenerated from banana protoplasts and transplanted in soil.
44 viability by FDA staining revealed that 90% of the protoplasts were viable as shown by yellow-green fluorescence. In co-culture with a suitable feeder, consisting of suspensions of very active cells of diploid banana, the protoplasts of triploid banana reformed the cell wall within 24 h, and underwent sustained divisions.
Plant regenerationfrom culturedprotoplasts After 3 days of co-culture with the feeder cells, first divisions occurred, leading to the formation of small clusters of 2-3 cells within 7 days. It has been shown earlier that the presence of active nurse cells was required to induce sustained divisions of protoplasts, which did not divide otherwise or rarely did (Megia et al. 1992). In this study, the plating efficiency ranged from 0.35 to 0.65% on day 10. The cell clusters of all sizes can be observed by the naked eye on day 14 (Fig. 1D). They underwent sustained cell divisions and gave rise to globular embryos of various sizes, and the older ones had a well-formed epidermis after day 50 (Fig. 1E). The somatic embryos developed a suspensor and a haustorium (Fig. 1F, G). No apparent callus phase was observed throughout the development of banana protoplastderived cell clusters into globular embryos. Further development into plants was obtained when protoplastderived embryos at globular and cotyledonous stages were incubated on regeneration medium. Only 3% of them gave rise to shoots and plantlets after 5 weeks of culture in 10 ~tM BAP (Fig. 1I). The frequency of plant regeneration was improved up to 10% by subculturing embryos in 2.2 lxlVI BAP and 2 ktM Picloram for one week, before transfer to MS medium containing 10 ~tM BAP. It indicated that most of the globular structures did not bear any shoot axis, since they were not yet bipolar, and a period of maturation was necessary to obtain a bipolar structure able to germinate. The rest of the primary embryos, cultured on medium with 10 ~tM BAP, underwent intensive direct secondary embryogenesis (Fig. 1H). This occurred from the upper layer of primary embryos, leading to the formation of multiple large somatic embryos and small single globular embryos. The multiple large somatic embryos continued to grow, and developed into shoots and plantlets with a frequency of up to 50% within 2 months, while only 7% of small embryos gave rise to plantlets on hormone-free medium. Secondary globular embryos were morphologically identical to primary embryos. Under appropriate conditions, they underwent either a further cycle of embryogenesis or plant development. Sufficiently developed shoots and plantlets were rooted on hormone-free MS medium before transfer to the greenhouse (Fig. 1J). As far as plant regeneration was concerned, the ontogenetic state of the explant source, and particularly embryogenic suspension cultures have proved determinant for successful protoplast cultures in monocotyledonous species (Vasil 1988). A reproducible protocol for culture of banana protoplasts has previously been developed (Megia et al. 1992). Plant regeneration has not, however, been achieved, probably due to the loss of totipotency of the old culture of diploid cell suspensions which had been used as protoplast source. In this study, the possibility that plantlet formation might have regenerated from the diploid cell suspensions used as feeder cells, was excluded because of
their inability to regenerate shoots. On the contrary, the use of freshly initiated embryogenic suspensions as protoplast source, displaying morphogenetic potential, has led to successful plant regeneration. Fast-growing and predominantly embryogenic suspensions cultures were highly desirable, as the system provided competent protoplasts (Vasil and Vasil 1987; Kyozuka et al. 1987; Funatsuki et al. 1992). However, it took a long time (at least 6 months) to establish such embryogenic suspensions of triploid banana, while at the same time suspensions tended to lose cell totipotency as they aged. In order to maintain morphogenic potential, it has been suggested either to routinely reinitiate cultures at intervals (Vasil 1988), or to use the technique of cryopreservation (Panis et al. 1990). In this study, the establishment of reproducible protoplast to plant system has successfully been achieved. However, the application of new techniques, including protoplast fusion and transformation, requires cultures of protoplasts from other genotypes of banana. Herein it has been shown that plants have directly been regenerated from cultured protoplasts via somatic embryogenesis, without passing through an apparent callus phase. Consequently, the morphogenetic events resulting in direct initiation of organized meristematic tissues may favour genetic stability (D'Amato 1975). Further morphological and cytological analyses, as well as field evaluation are needed for verifying the conformity of regenerated plants. Acknowledgements: This study was supported by a grant from EC/STD3. The authors wish to thank Dr. J.V. Escalant and the Laboratory Tropische Plantenteelt Leuven (K.U.L., Belgium) for kindly providing plant material, M. D. Froger for the photography.
References Bakry F (1984) Fruits 39:449-452 Chen WH, ZC Ku ZC (1985) J Agric Assoc China 129:56-57 Chu CC, Wang CC, Sun C.S. (1975) Sci Sinica 5:659-668 D'Amato F (1975) In: Frankel OH, Hawkes JG (eds) Crop Genetic Resources for Today and Tomorrow, Cambridge Unv. Press, pp 333348. Da Silva Conceicao A (1989) DEA G~n et S~lec Animale et Vegetale, Univ. Paris Sud, 14p Dhed'a D, Dmnortier F, Panis B, Vuylsteke D, De Langhe E (1991) Fruits 48:125-135 Escalant JV, Teisson C (1989) Plant Cell Rep 7:665-668 Escalant JV, Babeau J, Chatelet C, Teisson C (1992) In: Wills B (ed) Proc Inter Symp Genet. Improvement of bananas for resistance to diseases and pests, Montpellier - France, 7-9 Sept 1992 Funatsuki H, Ltrz H, Lazzeri PA (1992) Plant Sci 85:179-187 Galbraith DW (1981) Physiol Plant 53:111 - 116 Kao KN, Michaylnk MR (1975) Planta 126:105-110 Kyozuka J, Hayashi Y, Shimamoto K (1987) Mol Gen Genet 206: 408413 Lee L, Behroll RE, Grimes ND, Hodges (1989) Planta 178:325-333 Marroquin CG, Paduscheck C, Teisson C (1993) In Vitro Cell Dev Biol 29:43-46 Megia R, HaYcourR, Rossignol L, Sihachakr D (1992) Plant Sci 85:91-98 Morel G, Wetmore RH (1951) Am J Bot 38:141-143 Murashige T, Skoog F (1962) Physiol Plant 15:473-497 Novak FJ, Afza R, Van Duren M, Perea-DaUos M, Conger BV, Tang Xiao Lang (1989) Biotechnology 7:154-159 Panis BJ, Withers LA, De Langhe E (1990) Cryo Letters 11:337-350 Sannasgala AK (1989) PhD Thesis, Katholieke Univ Leuven, Belgium, 172 p Shillito RD, Carswell GK, Jonhson CM, Di Maio JJ, Harms CT (1989) Biotechnology 7:581-587 Vasil IK (1988) Biotechnology 6:397-402 Vasil V, Vasil IK (1980) Theor Appl Genet 56:97-99 Vasil V, Vasil IK (1987) Theor Appl Genet 73:793-798 Widhokn JM (1972) Stain Techno147:189-194
