Somatic Embryogenesis: A Model for Early Development in Higher Plants.

The ability to produce morphologically and developmentally normal embryos and, indeed, whole plants from undifferentiated somatic cells in culture, through the process of somatic embryogenesis, resides uniquely within the plant kingdom. Since the initial description of somatic embryo production from carrot callus cells more than 35 years ago (Steward et al., 1958), this unique developmental potential has been recognized both as an important pathway for the regeneration of plants from cell culture systems and as a potential model for studying early regulatory and morphogenetic events in plant embryogenesis. The last 5 to 10 years have witnessed an explosion in the number of species that can now be regenerated from cell culture into whole plants through somatic embryogenesis. The literature contains hundreds of references describing the specific manipulations required to effect somatic embryo development from a variety of agronomically and horticulturally important plants. Although this is obviously an extremely important application of the process of somatic embryogenesis, it is not the focus of this review. Rather, this review will focus on the use of somatic embryogenesis as a model system for understanding the regulation of gene expression required for the earliest developmental events in the life of a higher plant: the development of the fertilized zygote into a mature embryo. The events of fertilization and subsequent embryo development normally occur deep within maternal tissues. The early embryo is minute and is surrounded by both endosperm and maternal cells. Although the morphological description of embryo development has been extensively recorded through microscopy, molecular and biochemical analyses of early embryogenesis have been hampered significantly by this physical inaccessibility. As a consequence, we know very little about the genes that are necessary for early embryogenesis in higher plants and even less about their regulation. This is beginning to be remedied by recent intensive efforts to genetically identify genes required for early embryogenesis in model systems such as Arabidopsis and maize (see West and Harada, 1993, this issue); many interesting mutants have been identified that may provide entry points into molecular analyses of major morphogenetic events in embryogenesis. These analyses would be greatly enhanced by the availability of cell, tissue, and developmental stage-specific markers of important events in the differentiation of cells and the establishment of the major tissue systems of the plant, which occur early in embryogenesis. In addition, once genes have been identified that are essential for embryogenesis, the subsequent analysis of their regulation would be greatly facilitated by the availability of an appropriate in vitro model system that is not limited in tissue quantity or accessibility. The somatic embryo system represents just such a model system. This review will summarize the process of somatic embryogenesis and will address the strengths and limitations of somatic embryos as potential models for studying early events in plant embryo development.