dks8, a mutation specifically eliminating shoot formation during embryogenesis
--John D. Sollinger and Carol Rivin

From a group of maize embryo mutants generated in a Mutator-family transposon line, we have identified one that specifically eliminates the development of the shoot pole during embryogenesis, while the rest of the embryo and the endosperm undergo normal morphogenesis and maturation. The mutant, termed dks8 (defective kernel shootless-Mu8), is an excellent candidate for a pattern formation gene determining the earliest establishment of a functional shoot meristem. We are using descriptive and experimental embryology to characterize the role of this locus in development. Because dks8 appears to have been created by insertion of a Mu8 transposable element, we also have the opportunity to clone this gene and determine its product and expression.

We have examined the morphogenesis, growth and protein and lipid complement of homozygous dks8 mutant kernels over the course of embryogenesis and seed maturation. We find that dks8 kernels meet the criteria outlined by Jurgens et al. (1991) to distinguish mutants in pattern formation genes from other types of developmental abnormalities:

1. The morphological abnormality is specific. The morphological abnormalities of homozygous dks8 seeds are confined to components of the shoot system. Serial sections through mature mutant embryos show that there is no trace of leaf primordia, no coleoptile, no shoot vasculature, and no evidence for a shoot apical meristem. Neither do they reveal any sign of necrosis. The adaxial face of the scutellum is more circular than deltoid and is reduced in length by half, but we suspect that this is a secondary effect of the mutant condition, since the shoot normally acts both as a sink and as a source of developmental cues. Median longitudinal sections of mutant and wildtype embryos are shown in the accompanying figure.

2. The mutant phenotype deviates from wildtype at the time the pattern element is established. The shoot component can first be detected during normal embryogenesis with the formation of the coleoptilar ring. Using light microscopy and SEM, we examined developing dks8 mutant embryos and found that this feature never forms. Thus the mutant deviates from wildtype shoot development from a very early point in the establishment of that pattern element, suggesting a primary block to the specification of a shoot apical meristem, or to the function of that group of cells.

3. The mutant block should not interfere with the completion of other development events. The dks8 mutation has no effect on the capacity of the seed to complete other aspects of seed morphogenesis and differentiation. The mutant seeds are small, but otherwise morphologically normal and appropriately pigmented. They acquire a normal constellation of maturation proteins and storage lipids. They are desiccation tolerant and capable of germination of the primary root. Placed on a nutrient medium, this root will grow indefinitely, producing many root branches. However, there is no development of the secondary roots (products of the shoot system in maize). These features contrast sharply with the phenotype of other embryo mutants of maize, which are severely distorted in morphology, blocked from continued development, and unable to germinate.

Genetic studies of the dks8 mutation show that it is a single gene recessive condition. The dks8 kernels account for approximately 20% of the seeds on a self-fertilized ear, somewhat below the Mendelian expectation. We have found that the transmission of the dks8 allele through the pollen is equal to that of the wildtype locus. We expect, therefore, that the reduced percentage of dks8 homozygotes reflects a reduction in dks8 transmission through the ovule, a proposition we are now directly testing. It may be that the gene acts in the female gametophyte as well as in the embryo. Complementation tests conducted in our lab and with the help of Guy Farish, G at North Dakota have thus far shown that the dks8 mutation is not allelic to other mutants isolated by Bill Sheridan, WF and his co-workers that have abnormalities in embryonic shoot formation.

The dks8 mutation was isolated from an active Mutator transposon stock. We have conducted Southern blot analyses for co-segregation of various Mu elements with this mutant. In a survey of 70 individuals from several lineages and outcrosses, one Mu8 band consistently co-segregates with the dks8 allele. We are cloning the Mu8-containing fragment to map the dks8 locus and with the hope that we can determine the product of this gene. 

Return to the MNL 67 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage