SummaryWe implement a novel strategy for harnessing the power of high-copy transposons for functional analysis of the maize genome, and report behavioral features of the Mutator system in a uniform inbred background. The unique UniformMu population and database facilitate high-throughput molecular analysis of Mu-tagged mutants and gene knockouts. Key features of the population include: (i) high mutation frequencies (7% independent seed mutations) and moderation of copy number (approximately 57 total Mu elements; 1-2 MuDR copies per plant) were maintained by continuous back-crossing into a phenotypically uniform inbred background; (ii) a bz1-mum9 marker enabled selection of stable lines (loss of MuDR), inhibiting further transpositions in lines selected for molecular analysis; (iii) build-up of mutation load was prevented by screening Mu-active parents to exclude plants carrying pre-existing seed mutations. To create a database of genomic sequences flanking Mu insertions, selected mutant lines were analyzed by sequencing of MuTAIL PCR clone libraries. These sequences were annotated and clustered to facilitate bioinformatic subtraction of ancestral elements and identification of insertions unique to mutant lines. New insertions targeted low-copy, gene-rich sequences, and in silico mapping revealed a random distribution of insertions over the genome. Our results indicate that Mu populations differ markedly in the occurrence of Mu insertion hotspots and the frequency of suppressible mutations. We suggest that controlled MuDR copy number in UniformMu lines is a key determinant of these differences. The public database (http://uniformmu.org; http://endosperm.info) includes pedigree and phenotypic data for over 2000 independent seed mutants selected from a population of 31 548 F 2 lines and integrated with analyses of 34 255 MuTAIL sequences.

The Viviparous-1 [Vpl] gene is required for expression of the CI regulatory gene of the anthocyanin pathway in the developing maize seed. We show that VPl overexpression and the hormone, abscisic acid (ABA), activate a reporter gene driven by the CI promoter in maize protoplasts. Cis-acting sequences essential for these responses were localized. Mutation of a conserved sequence in the CI promoter abolishes both ABA regulation and VPl trans-activation. An adjacent 5-bp deletion blocks ABA regulation but not VPl trans-activation. The latter mutant reconstructs the promoter of cl-p, an allele that is expressed during seed germination but not during seed maturation. We suggest that VPl activates CI specifically during maturation by interacting with one or more ABA-regulated transcription factors.

The key regulatory step in starch biosynthesis is catalyzed by the tetrameric enzyme ADP-glucose pyrophosphorylase (AGPase). In leaf and storage tissue, the enzyme catalyzes the synthesis of ADP-glucose from glucose-1-phosphate and ATP. Using heterologous probes from maize, two sets (B and S) of cDNA clones encoding potato AGPase were isolated from a tuberspecific cDNA library. Sequence analysis revealed homology to other plant and bacterial sequences. Transcript sizes are 1.9 kb (AGPase B) and 2.1 kb (AGPase S). Northern blot experiments show that the two genes differ in their expression patterns in different organs. Furthermore, one of the genes (AGPase S) is strongly inducible by metabolizable carbohydrates (e.g. sucrose) at the RNA level. The accumulation of AGPase S mRNA was always found to be accompanied by an increase in starch content. This suggests a link between AGPase S expression and the status of a tissue as either a sink for or a source of carbohydrates. By contrast, expression of AGPase B is much less variable under various experimental conditions.

Mutation at the shrunken-2 (Sh2) locus of maize, a gene described more than 40 years ago, greatly reduces starch levels in the endosperm through its effect on the starch synthetic enzyme ADP-glucose pyrophosphorylase, an enzyme thought to be regulatory in this biosynthetic pathway. Although our previous work has suggested that Sh2 is a structural gene for this enzyme, we have also reported data compatible with Sh2 acting post-transcriptionally.In this study, we took advantage of a transposable element-induced Sh2 allele, its progenitor, and revertants to identify a clone for this locus. Although the cloning and identification were done independently of any knowledge concerning the product of this gene, examination of the deduced amino acid sequence revealed much similarity to known ADP-glucose pyrophosphorylase subunits of plants and bacteria, including regions involved in substrate binding and activator binding. Little sequence similarity, however, was found at the DNA level. These obsenrations provide direct evidence that Sh2 encodes a subunit for endosperm ADP-glucose pyrophosphorylase. Analysis of several phenotypically wild-type alleles arising from a mutable sh2-Ds allele revealed one unexpected case in which DNA sequences of Sh2 were rearranged in comparison with the progenitor Sh2. In contrast to wild type, the Dsinduced sh2 allele conditions at least two transcripts in the endosperm.