Gene expression in response to abscisic acid and osmotic stress.

Abscisic acid (ABA) was discovered in the 1950s to be a phytohormone affecting leaf abscision and bud dormancy. It was soon characterized as a sesquiterpene derived from mevalonate although certain steps of its biosynthesis in plants are still unknown (Li and Walton, 1987; Zeevaart and Creelman, 1988). Continuing work on ABA has shown that it mediates various developmental and physiological processes that affect the agronomic performance of crop plants (Austin et al., 1982; Ramagopal, 1987). These processes include embryo maturation and germination as well as the response of vegetative tissues to osmotic stress (Singh et al., 1987; Zeevaart and Creelman, 1988). ABA levels increase in tissues subjected to osmotic stress by desiccation, salt, or cold (Henson, 1984; Mohapatra et al., 1988). Under these conditions, specific genes are expressed.that can also be induced in unstressed tissues by the application of exogenous ABA (Singh et al., 1987; Gomez et al., 1988; Mundy and Chua, 1988). Some of these genes are also expressed during the normal embryogenic program when seeds desiccate and embryos become dormant (Dure et al., 1981). Although different sets of ABA-responsive genes exhibit different patterns of developmental and tissue-specific expression, some of them appear to be part of a general reaction to osmotic stress. This system is a normal part of the embryogenic program but is inducible in vegetative tissues at other times in the plant life cycle. Severa1 ABA-responsive genes have now been isolated (Baker et al., 1988; Gomez et al., 1988; Marcotte et al., 1988; Mundy and Chua, 1988; Vilardell et al., 1990; Yamaguchi-Shinozaki et al., 1990). A major goal of the research discussed below is to understand the role these genes play in osmotic stress and desiccation tolerance. ABA-responsive genes are also being studied as tools to develop molecular models of ABA action. So far, work on animal hormones has defined two general mechanisms of action: (1) regulation of transcription factors by steroid hormone binding (Beato, 1989), and (2) activation of regulatory factors via “second messenger” pathways (Deutsch et al., 1988). In contrast, models of ABA action are incomplete, in part because ABA receptors have yet