American Journal of Botany 90(7): 973979. 2003.

COMPARATIVE ANATOMY AND MORPHOLOGY OF VITIS VINIFERA (VITACEAE) SOMATIC EMBRYOS FROM SOLID- AND LIQUID-CULTURE-DERIVED
PROEMBRYOGENIC MASSES1

S. JAYASANKAR,2 BHASKAR R. BONDADA,3 ZHIJIAN LI, AND D. J. GRAY4

Mid-Florida Research and Education Center, Institute of Food Agricultural Sciences, University of Florida, 2725 Binion Road, Apopka, Florida 32703-8504 USA Ontogeny of somatic embryos of grapevine (Vitis vinifera) produced from solid- and liquid-culture-derived proembryogenic masses (PEM) was compared using light and scanning electron microscopy. Somatic embryos produced from solid-medium-derived PEM (SPEM) had large cotyledons, little or no visible suspensor structure, and a relatively undeveloped concave shoot apical meristem, whereas those from liquid-medium-derived PEM (LPEM) had smaller cotyledons, a distinct suspensor, and a at-to-convex shoot apical meristem. The convex shoot apical meristem in LPEM-derived somatic embryos formed as early as the heart stage of development; it was 46 cell layers deep and rich in protein. Suspensors persisted in fully developed and mature LPEM-derived somatic embryos. The SPEM-derived somatic embryos exhibited dormancy, as do mature zygotic embryos, which also have a rudimentary suspensor, whereas LPEM-derived embryos were not dormant. We hypothesize that the presence of a persistent suspensor in LPEM-derived somatic embryos modulates development, ultimately resulting in rapid germination and a high plant-regeneration rate. Key words: cell culture; embryogenesis; grapevine; somatic embryogenesis; Vitaceae; Vitis vinifera.

Although somatic and zygotic embryos are nearly identical in structural and functional characteristics (Goldberg et al., 1989, 1994; Gray and Purohit, 1991a), the ontogeny of somatic embryos tends to be more variable (Litz and Gray, 1992). Somatic embryos of most species, including grapevine (Vitis spp.), tend to exhibit several typical morphological abnormalities such as variation in shape, size, and cotyledon number (Goebel-Tourand et al., 1993; Gray, 1995; Jayasankar et al., 2002). In most species, somatic embryos are larger than zygotic embryos, and their regeneration rates are lower. For instance, in Vitis rupestris Scheele, only 3% of somatic embryos were capable of developing into complete plants, although 27% had shoot and root apices (Faure, 1990). Grapevine somatic embryogenesis, rst reported by Mullins and Srinivasan (1976), has become commonplace for several genotypes (Gray, 1995). In most reports, initiation and maintenance of embryogenic cultures were accomplished by growth on solidied medium, and plant-regeneration rates were often low ( 20%). Recently, we demonstrated an embryogenic liquid culture system in which somatic embryos differ morphologically from their solid-medium-derived counterparts and exhibit a higher plant-regeneration rate ( 60%) (Jayasankar et al., 1999). In the present study, we show the
1 Manuscript received 3 December 2002; revision accepted 13 February 2003. The authors thank Ms. Diann Achor, CREC, IFAS, University of Florida, Lake Alfred for help in scanning electron microscopy. This research was supported by the Florida Agricultural Experiment Station and a grant from The Florida Department of Agriculture and Consumer Services Viticulture Trust Fund, Tallahassee, Florida and approved for publication as Journal Series No. R-09251. 2 Present address: Department of Plant Agriculture, Ontario Agricultural College, University of Guelph, Vineland Campus, 4890 Victoria Ave. N., P.O. Box 7000, Vineland Station, Ontario, Canada L0R 2E0. 3 Present address: University of California, Department of Viticulture and Enology, One Shields Avenue, Davis, California 95616 USA. 4 Author for reprint requests (e-mail: djg@mail.ifas.u.edu).

ontogenic pattern to be set very early (i.e., in proembryogenic masses [PEM]), because PEM initiated in solid or liquid culture systems then plated onto solid medium produced somatic embryos with characteristic morphological and developmental differences. Anatomical and morphological studies have proven useful in understanding somatic and zygotic embryo development (Altamura et al., 1992; Gray, 1995; Faure et al., 1996a, b). Hence, we employ the same approach to compare somatic embryogenesis from liquid- and solid-medium-derived PEM (LPEM and SPEM, respectively) to investigate reasons for differential ontogeny and plant-regeneration efciency. MATERIALS AND METHODS
Culture maintenanceEmbryogenic cultures of Vitis vinifera L. Chardonnay (Clone 15) and Thompson Seedless were induced from anthers and leaves, respectively, and maintained either on solid medium as described by Gray (1995) or in liquid culture and maintained as PEM as described by Jayasankar et al. (1999). To produce somatic embryos for study, PEM from solid and liquid culture systems were plated onto solidied X6 medium as previously described (Li et al., 2001). Zygotic embryos also were excised from fully mature seeds of Chardonnay for comparative study. Embryos at various developmental stages were selected using a stereomicroscope and xed immediately for light and scanning electron microscopy. Scanning electron microscopy (SEM)All xation and rinse solutions were buffered to pH 6.8 with 0.05 mol/L Sorensons phosphate buffer and kept at 4 C. Unless otherwise specied, all xation and dehydration steps were done on ice. Embryos were xed in cold 3% gluteraldehyde and incubated overnight at 4 C. After rinsing with buffer (three times), embryos were xed in 1% osmium tetroxide for 30 min and then rinsed three times. Embryos were then dehydrated in a graded ethanol series (15% increments every 30 min). After two changes (15 min each) of 100% ethanol, embryos were subjected to critical-point drying, mounted on stubs, sputtered with gold-palladium, and immediately observed with a Hitachi model S530 SEM.

973

974

AMERICAN JOURNAL

OF

BOTANY

[Vol. 90

Light microscopyEmbryos were xed in glutaraldehyde and serially dehydrated in ethanol as described. A trace amount of safranin was added to the rst 100% ethanol rinse, and specimens were incubated at room temperature for 15 min to introduce color into the normally translucent embryos for visualization in the resin for micro-orientating and sectioning. After two quick rinses with 100% ethanol to remove excess safranin, embryos were transferred to cold 100% ethanol for another 15 min. Dehydrated embryos were embedded in at molds in JB4-Plus plastic resin (Polysciences, Pennsylvania, USA), as per manufacturers instructions. The embryos were carefully oriented in the molds to facilitate subsequent median longitudinal sectioning. Small blocks of resin, approximately 14 mm3, each containing an embryo or cluster of embryos, were excised from the molds and attached to the end of solidied resin blanks with Superglue in an orientation for cutting of median longitudinal sections. Specimens were reoriented several times during facing and sectioning to obtain median longitudinal sections. Sections approximately 5 m thick were cut with glass knives and secured to slides by heating at 70 C. The sections were stained with periodic acid-Schiffs reagent (PAS) to contrast polysaccharides (e.g., cell walls and amyloplasts) and counter stained with naphthol blue black to visualize proteins.

meristem was better developed than that of SPEM-derived embryos, being at-to-convex in cross section and composed of four or more cell layers (Figs. 12, 13). The root apical meristem and embryonic vasculature did not differ substantially from SPEM-derived somatic embryos (Figs. 12, 14). Development of somatic embryosFigures 1524 reconstruct the developmental sequence of LPEM-derived somatic embryos, based upon individually selected specimens. Early stages of embryogenesis occurred rapidly from LPEM. Small globular somatic embryos were observed within 2 wk after plating LPEM onto solid medium. As embryos reached the globular stage, they protruded above subtending tissue on an elongate suspensor (Fig. 15), and the distinction between the suspensor and the embryo proper became clear. As the embryo body continued to enlarge, it initially maintained a radial symmetry (Figs. 1518). Embryos up to this stage were approximately 500 m or less in length, of which more than 60% was the suspensor. Assumption of bilateral symmetry occurred as early as 3 wk after plating LPEM onto solid medium, as evidenced by differential enlargement from two distinct sites of accelerated growth; this signaled the start of cotyledon development and resulted in the appearance of a concave dimple in the distal end of the embryo (Figs. 19, 20, arrows). Further cell divisions and enlargement led to elongation of the embryo body and early denition of the cotyledons (Fig. 20, arrow). The cotyledonary poles rapidly outgrew the central region, leaving a distinct notch between them (Fig. 21, arrow). This corresponded to a typical heart stage of development. Also, at this stage, the dimple became convex in many LPEM-derived somatic embryos from cell divisions in the shoot apical meristem (Fig. 22, arrow). The cotyledons differentiated further, completely enveloping the shoot apical meristem, and the hypocotyl continued to enlarge as embryos reached the torpedo stage of development. Embryos at this stage were 1.0 2.5 mm in length. Within 5 wk of plating, well-dened, morphologically correct dicotyledonous somatic embryos were apparent (Fig. 23). Mature LPEM-derived somatic embryos (Fig. 24) typically differed from SPEM-derived somatic embryos (Fig. 25) by their large, persistent suspensors and smaller cotyledons. The SPEM-derived somatic embryos were nearly identical in morphology to zygotic embryos in that they both had heart-shaped cotyledons typical of grapevine; however, somatic embryos were noticeably larger than zygotic embryos, which tended to be attened (compare Figs. 25, 26). As with some SPEMderived somatic embryos, zygotic embryos had a rudimentary suspensor (Fig. 26). The relative length of hypocotyl-to-cotyledon varied among the three embryo types. In SPEM-derived somatic embryos, the cotyledons were longer than the hypocotyls. In LPEM-derived somatic embryos, the hypocotyls were longer than the cotyledons, whereas in zygotic embryos the lengths were approximately equal. DISCUSSION In SPEM-derived culture material, typical development of grapevine somatic embryos was nearly identical to that of zygotic embryos (Gray, 1995). Somatic embryos have more morphological variation, such as pluricotyly, than zygotic embryos, which are smaller and attened, probably from differences in the in vitro and in vivo (seed) environments. In the seed, the developing zygotic embryo is physically compressed (Gray

RESULTS Embryogenic culturesRegardless of solid or liquid medium origin, an embryogenic culture typical of grapevine developed when PEM were transferred to and grown on solidied X6 medium. Because embryogenic cultures grew nonsynchronously, somatic embryos in different stages of development were present at any given time (Figs. 1, 3, 10). Somatic embryos were white and opaque (Fig. 1). Despite the general similarity among embryogenic cultures, SPEM- and LPEMderived somatic embryos differed in their structural and functional characteristics. SPEM-derived somatic embryosThe SPEM-derived somatic embryos typically developed with enlarged (often supernumerary) cotyledons (Figs. 1, 5). When isolated from the underlying cell mass, a narrowed point of attachment reminiscent of a suspensor was apparent (Fig. 2). As a result of this limited attachment, such embryos were sessile with respect to underlying tissue (Fig. 3). Repetitive embryogenesis occurred in the immediate subtending tissue (Fig. 4, arrows), resulting in typical groupings of somatic embryos, which often appeared to be fused to each other at their bases, but were, in fact, attached only through this embryogenic tissue. The shoot apical meristem of solid-medium-derived somatic embryos typically was reduced in size, concave in cross section, and composed of 23 cell layers (Fig. 6). The root apical meristem appeared to be well developed, with typical embryonic vasculature spanning from proximal to distal regions of the embryo body (Figs. 5, 7). LPEM-derived somatic embryosThe LPEM-derived somatic embryos differed from their SPEM-derived counterparts in that they had an enlarged suspensor apparatus, which was apparent from the earliest stages of development (Fig. 8) and persistent through maturity (Figs. 912). Such embryos developed without attachment/fusion to each other and often became perched well above subtending tissue (Figs. 8, 9, 10). The suspensor was multiseriate, often exceeding seven cell layers in width (Fig. 11). In early stages, a clear demarcation between the suspensor region and embryo proper could not be established. Like SPEM-derived somatic embryos, LPEM-derived embryos also often had supernumerary cotyledons, but they were distinctly smaller in size (Fig. 12). The shoot apical

July 2003]

JAYASANKAR

ET AL.SOMATIC EMBRYO ONTOGENY IN GRAPEVINE

975

Figs. 17. The solid-medium-derived somatic embryogenesis in grapevine. 1. Embryogenic culture maintained on solid medium with white, opaque somatic embryos. Bar 520 m. 2. Isolated embryo with well-dened cotyledons and rudimentary suspensor. Bar 125 m. 3. Asynchronous development of somatic embryos. Note sessile nature of embryos relative to underlying tissue and pluricotyly of large embryo in upper center. Bar 340 m. 4. Attachment of somatic embryo to embryogenic tissue. Note secondary embryogenesis (arrows) and lack of dened suspensor. Bar 135 m. 5. Mature somatic embryo with large cotyledons and well-developed embryonic vasculature. Bar 275 m. 6. Concave shoot apical meristem typical of solid-medium-derived somatic embryos. Bar 65 m. 7. Base of mature somatic embryo detailing vasculature and root apical meristem. Bar 90 m.

976

AMERICAN JOURNAL

OF

BOTANY

[Vol. 90

Figs. 814. The liquid-medium-derived somatic embryogenesis in grapevine. 8. Somatic embryos with prominent suspensors originating from liquid-mediumderived proembryogenic masses (LPEM), transferred to solid medium for embryogenesis. Bar 200 m. 9. Pluricotyledonous somatic embryo connected to PEM by a suspensor. Bar 260 m. 10. Asynchronous development of somatic embryos. Note that embryos of all stages present exhibit elongated suspensors. Bar 400 m. 11. Attachment of heart-stage embryo to suspensor. Note that the demarcation between suspensor and embryo proper is not evident at this stage. Bar 185 m. 12. Mature somatic embryo with small cotyledons, detailing embryonic vasculature. Bar 160 m. 13. Convex shoot apical meristem typical of liquid-medium-derived somatic embryos. Bar 58 m. 14. Base of mature somatic embryo showing presence of suspensor. Bar 50 m.

July 2003]

JAYASANKAR

ET AL.SOMATIC EMBRYO ONTOGENY IN GRAPEVINE

977

Figs. 1526. Developmental morphology of liquid-medium-derived somatic embryos (1524) compared with solid-medium-derived somatic embryo (25) and zygotic embryo (26). 15. Initially, the suspensor (Su) is the most prominent part of the developing embryo apparatus. 16. As the embryo body (Em) begins to differentiate, it becomes more obvious. 1718. Continued development results in a late-stage, globular embryo. 19. A concave dimple (arrow) in the distal end of the embryo is the rst sign of cotyledons and development of shoot apical meristem. 20. The embryo body begins to elongate and the dimpled area becomes more prominent (arrow). 21. Cotyledon enlargement (arrow) results in a distinct cleft. 22. The dimpled area (arrow) becomes convex as the shoot apical meristem differentiates. 23. Continued growth results in an elongated embryo with cotyledons that enclose the shoot apical meristem. 24. A mature liquid-medium-derived somatic embryo exhibits smaller cotyledons and a persistent suspensor (Su), when compared to solid-medium-derived somatic embryo (25) or a zygotic embryo (26). The cotyledons of solid-medium-derived somatic embryo and zygotic embryo are typically heart-shaped; however, the zygotic embryo is smaller and laterally attened in comparison. Bars 500 m.

and Purohit, 1991a). Developmentally, both SPEM-derived somatic embryos and their zygotic counterparts exhibit physiological dormancy; they do not germinate to undergo plant development without a dormancy-breaking pretreatment. It was surprising to discover that two morphologically distinct types of grapevine somatic embryos could be produced on solid medium; the only apparent difference was the initial culture environment in which the PEM were grown. When compared to somatic embryos from SPEM, somatic embryos from LPEM possessed a large, persistent suspensor, smaller cotyledons, and a more anatomically dened shoot apical meristem. Developmentally, LPEM-derived somatic embryos also

differed in that they did not exhibit dormancy (i.e., they germinated readily without a pretreatment to break dormancy) and produced a higher percentage of plants compared to SPEM-derived somatic embryos ( 60% vs. 20%). It is tempting to link structural differences to the observed developmental and physiological differences between solid- and liquid-medium-derived somatic embryos. In contrast to the persistent and robust suspensor of LPEMderived somatic embryos, the typical suspensor in zygotic embryos is programmed to die when the embryo proper reaches the torpedo stage (Yeung and Meinke, 1993). The suspensors in LPEM-derived somatic embryos are often long, multiser-

978

AMERICAN JOURNAL

OF

BOTANY

[Vol. 90

iate, and relatively massive. Their cytoplasm-rich cells suggest that they are metabolically active, perhaps in supplying nutrients or modulating the hormonal balance in the developing somatic embryos (Souter and Lindsey, 2000). In contrast, the suspensors in mature zygotic embryos, when detected, are only a few cells long. The cells are often highly vacuolated, suggesting that they are not cytologically active. The suspensor originally was thought to be an embryonic accessory that was necessary for positioning the developing embryo in the embryo sac. However, studies with common bean (Phaseolus vulgaris L.) have shown that it is necessary for active protein synthesis in developing zygotic embryos (Brady and Walthall, 1985). Later, Nagl et al. (1991) demonstrated the synthesis of storage proteins in the suspensor cells of beans. We earlier observed that LPEM-derived somatic embryos germinate precociously without undergoing the dormancy typical of grapevine zygotic embryos (Jayasankar et al., 1999), whereas somatic embryos derived from SPEM exhibit dormancy (Gray, 1989; Gray and Purohit, 1991a, b), and require a pretreatment to germinate (Gray and Mortensen, 1987; Gray and Purohit, 1991b). Such dormancy in grapevine somatic embryos is attributed to ABA accumulation, which typically reaches a peak during maturation (Rajasekaran et al., 1982). However, exogenous GA3-induced grapevine somatic embryo germination, and the concentration of GA-like compounds also increased during cold stratication (Takeno et al., 1983; Pearce et al., 1987). The persistent suspensor in LPEM-derived somatic embryos may be a reason for the lack of dormancy and concomitant precocious germination. Suspensors also are a site of gibberellin synthesis. Very high levels of GA-like substances have been found in suspensors of Trapeolum (Picciarelli et al., 1984), Cytisus (Picciarelli et al., 1991), and Phaseolus (Piaggesi et al., 1989). Studies in P. coccineus L. have shown that GA moves from suspensor to the embryo proper during embryo maturation (Alpi et al., 1975, 1979). We hypothesize that gibberellins are produced abundantly and continuously in the suspensors of LPEM-derived somatic embryos and then are transported to the embryo proper, leading to precocious germination. Because certain nutrients and growth regulators are specifically produced only in the suspensor, lack of an adequate suspensor may deprive somatic embryos of growth factors needed for proper shoot apical meristem development. This lack is suggested by the poor apical meristem development exhibited by SPEM-derived somatic embryos, which have premature vacuolation in the meristematic region. Embryos with such a defective meristem are not likely to develop into a complete plant. Premature vacuolation also has been implicated in the failure to develop a functional meristem in carrot (Nickle and Yeung, 1993) and canola (Yeung et al., 1996). Such poor development of the meristematic region in many SPEM-derived somatic embryos may be a factor leading to low efciency of plant regeneration. Plant regeneration using somatic embryogenesis is necessary to produce novel genotypes after genetic manipulation. This necessity is particularly evident considering that many plant transformation systems, including grapevine (e.g., Li et al., 2001), employ embryogenic culture systems as a source of target cells. However, little attention has been paid to the developmental processes leading to regeneration of a whole plant from a single somatic cell. Yeung and Stasolla (2000) proposed that the capacity of a somatic embryo to regenerate into a plant largely depends on the quality of its shoot apical

meristem. Our research extends that viewpoint to suggest that proper development of the shoot apical meristem depends, in turn, on the presence of certain embryonic organs, particularly the suspensor. LITERATURE CITED
ALPI, A., R. LORENZI, P. G. CIONINI, A. BENNICI, AND F. DAMATO. 1979. Identication of gibberellin A1 in the embryo suspensor of Phaseolus coccineus. Planta 147: 225228. ALPI, A., F. TOGNONI, AND F. DAMATO. 1975. Growth regulator levels in embryo and suspensor of Phaseolus coccineus at two stages of development. Planta 127: 153162. ALTAMURA, M. M., A. CERSOSIMO, C. MAJOLI, AND M. CRESPAN. 1992. Histological study of embryogenesis and organogenesis from anthers of Vitis rupestris du Lot cultured in vitro. Protoplasma 171: 134141. BRADY, T., AND E. D. WALTHALL. 1985. The effect of the suspensor and gibberellic acid on Phaseolus vulgaris embryo protein content. Developmental Biology 107: 531536. FAURE, O. 1990. Embryons somatiques de Vitis rupestis et embryons zygotiques de Vitis sp.: morphologie, histology, histochemie et developpement. Canadian Journal of Botany 68: 23052315. FAURE, O., J. AAROUF, AND A. NOUGAREDE. 1996a. Ontogenesis, differentiation and precocious germination in anther-derived somatic embryos of grapevine (Vitis vinifera L.): proembryogenesis. Annals of Botany 78: 2328. FAURE, O., J. AAROUF, AND A. NOUGAREDE. 1996b. Ontogenesis, differentiation and precocious germination in anther-derived somatic embryos of grapevine (Vitis vinifera L.): embryonic ontogenesis. Annals of Botany 78: 2937. GOEBEL-TOURAND, I., M.- C. MAURO, L. SOSSOUNTZOV, E. MIGIANIAC, AND A. DELOIRE. 1993. Arrest of somatic embryo development in grapevine: histological characterization and the effect of ABA, BAP and zeatin in stimulating plantlet development. Plant Cell, Tissue and Organ Culture 33: 91103. GOLDBERG, R. B., S. J. BARKER, AND L. PEREZ-GRAU. 1989. Regulation of gene expression during plant embryogenesis. Cell 56: 149160. GOLDBERG, R. B., G. PAIVA, AND R. YADEGIRI. 1994. Plant embryogenesis: zygote to seed. Science 266: 605614. GRAY, D. J. 1989. Effects of dehydration and exogenous growth regulators on dormancy, quiescence and germination of grape somatic embryos. In Vitro Cellular and Devlopmental Biology-Plant 25: 11731178. GRAY, D. J. 1995. Somatic embryogenesis in grape. In S. M. Jain, P. K. Gupta, and R. J. Newton [eds.], Somatic embryogenesis in woody plants, vol. 2, 191218. Kluwer Academic, Dordrecht, Netherlands. GRAY, D. J., AND J. A. MORTENSEN. 1987. Initiation and maintenance of long term somatic embryogenesis from anthers and ovaries of Vitis longii Microsperma. Plant Cell Tissue and Organ Culture 9: 7380. GRAY, D. J., AND A. PUROHIT. 1991a. Somatic embryogenesis and development of synthetic seed technology. Critical Reviews in Plant Science 10: 3361. GRAY, D. J., AND A. PUROHIT. 1991b. Quiescence and dormancy in somatic embryos. In Y. P. S. Bajaj [ed.], Biotechnology in agriculture and forestry, vol. 17, 382394. Springer-Verlag, Berlin, Germany. JAYASANKAR, S., B. R. BONDADA, Z. LI, AND D. J. GRAY. 2002. A unique morphotype of grapevine somatic embryos exhibits accelerated germination and early plant development. Plant Cell Reports 20: 907911. JAYASANKAR, S., D. J. GRAY, AND R. E. LITZ. 1999. High frequency somatic embryogenesis and plant regeneration using suspension culture in grapevine Vitis vinifera L.). Plant Cell Reports 18: 533537. LI, Z., S. JAYASANKAR, AND D. J. GRAY. 2001. Expression of a bifunctional green uorescent protein (GFP) fusion marker under the control of three constitutive promoters and enhanced derivatives in transgenic grape (Vitis vinifera). Plant Science 160: 877887. LITZ, R. E., AND D. J. GRAY. 1992. Organogenesis and somatic embryogenesis. In F. A. Hammerschlag and R. E. Litz [eds.], Biotechnology of perennial fruit crops, 334. CAB International, Wallingford, UK. MULLINS, M. G., AND C. SRINIVASAN. 1976. Somatic embryos and plantlets from an ancient clone of grapevine (cv. Cabernet-Sauvignon) by apomixis in vitro. Journal of Experimental Botany 27: 10221030. NAGL, W., T. SEELIGER, AND A. HARTMANN. 1991. Immunocytochemical localization of the storage protein phaseolin in the embryo suspensor of

July 2003]

JAYASANKAR

ET AL.SOMATIC EMBRYO ONTOGENY IN GRAPEVINE

979

Phaseolus coccineus. European Journal of Basic and Applied Histochemistry 35: 175183. NICKLE, T. C., AND E. C. YEUNG. 1993. Failure to establish a functional shoot meristem may be a cause of conversion failure in somatic embryos of Daucus carota (Apiaceae). American Journal of Botany 80: 12841291. PEARCE, D., R. P. PHARIS, K. RAJASEKARAN, AND M. G. MULLINS. 1987. Effects of chilling and ABA on [3H] gibberellin A4 metabolism in somatic embryos of grape. Plant Physiology 80: 381385. PIAGGESI, A., P. PICCIARELLI, R. LORENZI, AND A. ALPI. 1989. Gibberellins in embryo-suspensor of Phaseolus coccineus seeds at the heart stage of embryo development. Plant Physiology 91: 362366. PICCIARELLI, P., A. ALPI, L. PISTELLI, AND M. SCALET. 1984. Gibberellinlike activity in suspensors of Tropaeolum majus and Cytisus laburnum. Planta 162: 566568. PICCIARELLI, P., A. PIAGGESI, AND A. ALPI. 1991. Gibberellins in suspensor, embryo and endosperm of developing seeds of Cytisus laburnum. Phytochemistry 30: 17891792. RAJASEKARAN, K., J. VINE, AND M. G. MULLINS. 1982. Dormancy in somatic

embryos and seeds of Vitis: changes in endogenous abscisic acid during embryogeny and germination. Planta 154: 139144. SOUTER, M., AND K. LINDSEY. 2000. Polarity and signaling in plant embryogenesis. Journal of Experimental Botany 51: 971983. TAKENO, K., M. KOSHIOKA, R. P. PHARIS, K. RAJASEKARAN, AND M. G. MULLINS. 1983. Endogenous gibberellin-like substances in somatic embryos of grape (Vitis vinifera Vitis rupestris) in relation to embryogenesis and the chilling requirement for subsequent development of mature embryos. Plant Physiology 73: 803808. YEUNG, E. C., AND D. W. MEINKE. 1993. Embryogenesis in angiosperms: development of the suspensor. Plant Cell 5: 13711381. YEUNG, E. C., M. H. RAHMAN, AND T. A. THORPE. 1996. Comparative development of zygotic and microspore-derived embryos in Brassica napus L. Cv. Topas. I. Histodifferentiation. International Journal of Plant Science 157: 2739. YEUNG, E. C., AND C. STASOLLA. 2000. Somatic embryogenesisapical meristems and embryo conversion. Korean Journal of Plant Tissue Culture 4: 299307.