GA and the germination of embryos in vivo and in culture
--Connie White and Carol Rivin

The phytohormone abscisic acid (ABA) has been shown to be an important regulator of maize embryo maturation. Excised immature embryos are inhibited from germination and synthesize a complement of embryo proteins in response to culture with ABA. Certain viviparous mutants appear to undergo precocious germination as a result of diminished ABA levels. Because of the strong antagonism between ABA and gibberellic acid (GA) documented in other monocot systems, and because of the known role for GA in the germination of other cereals, we examined the effects of GA and GA biosynthesis inhibitors on germination processes of both wildtype and viviparous embryos in culture and in vivo.

We assayed germinability of embryos excised at stages 2, 3, and 4. Stage 2 and 3 wildtype embryos germinate readily in culture, while stage 4 embryos are quiescent. We asked whether disruption of GA synthesis would alter germination. Germination of stage 2 embryos was relatively unaffected by the inhibitors ancymidol and paclobutrazol, while germination of stage 3 embryos was suppressed in the presence of these inhibitors. Inhibitor-mediated suppression was alleviated by the addition of exogenous GA3. Stage 4 embryos cultured with exogenous GA3 germinated readily. Our simplest interpretation of these results is that the presence of GA is required for in vitro germination and that stage-specific variation in germinability and/or sensitivity to GA and GA inhibitors reflects variation in endogenous ABA and GA levels or synthetic capacity.

In addition to affecting germination, these growth regulators also influenced the levels of anthocyanin synthesized in wildtype embryos. Anthocyanin synthesis was enhanced by either ancymidol or paclobutrazol and antagonized by exogenous GA3. We found that the formation of anthocyanins in cultured embryos required the presence of an R allele conditioning embryo pigmentation in vivo.

We tested the effects of altered GA metabolism in two viviparous mutants both in vitro and in vivo. vp1 (ABA insensitive) and vp5 (ABA deficient) embryos were excised at various stages and cultured in the presence and absence of GA inhibitors. Germinability of either genotype was relatively unaffected by any treatment. We disrupted GA synthesis in vivo by biochemical and genetic means. In the first protocol, one side of vp1 ears were treated with either of two commercial preparations of inhibitors, "A-rest" (ancymidol plus surfactant--Dow Elanco) or "Bonzi" (paclobutrazol plus surfactant--ICI UniRoyal) at either 10 or 100 micromolar concentrations at 5 day intervals starting with 6 days after pollination and ending at 26 days after pollination. Other ears were treated with 10 or 100 micromolar GA3 alone or in combination with 100 micromolar paclobutrazol. The treated ears were examined 50 days after pollination. Compared to the untreated half and to control ears, none of the treatments induced a notable difference in degree or extent of vivipary. Wildtype ears in the same genetic background (W22) were treated in a like manner and similarly showed no altered effects either in kernel maturation or germinability after drying. Our genetic protocol involved planting the F2 generation of a cross between vp1 plants and dwarf1 (d1) plants. Dwarf plants were both selfed and outcrossed to standard vp1 stock to test for expression and presence of the vp1 allele. Self-pollinated ears of d1/d1, vp1/+ plants segregated for vivipary in the expected 3:1 ratio. The extent of vivipary was not dramatically different between selfed ears and ears pollinated by a standard vp1 stock. Shoot growth was slightly diminished on ears whose mother plant was d1/d1. Similar experiments are currently being undertaken on vp5 mutants.

Our results indicate that while de novo GA synthesis is required for germination of immature wildtype embryos, germination of cultured vp1 or vp5 embryos does not depend on the synthesis of this hormone. In addition, expression of vivipary by vp1 embryos seems unaffected by diminished GA levels in vivo. These differences in sensitivity may indicate that germination of cultured wildtype embryos proceeds along an entirely different pathway than that which occurs with viviparous embryos or alternatively, that our crude methods are obscuring some delicate and perhaps transient GA-mediated regulation of vivipary. Since germination may involve a balance of ABA and GA, experiments to assay the effects of lowered GA levels on vp5 embryos in vivo should enable us to distinguish between these two possibilities. 

Please Note: Notes submitted to the Maize Genetics Cooperation Newsletter may be cited only with consent of the authors

Return to the MNL 67 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page