SummaryTo devise a method for function-based gene isolation and characterization in barley, we created a plasmid containing the maize Activator (Ac) transposase (AcTPase) gene and a negative selection gene, codA, and a plasmid containing Dissociation (Ds) inverted-repeat ends surrounding the selectable herbicide resistance gene, bar. These plasmids were used to stably transform barley (Hordeum vulgare). In vitro assays, utilizing a Ds-interrupted uidA reporter gene, were used to demonstrate high-frequency excisions of Ds when the uidA construct was introduced transiently into stably transformed, AcTPaseexpressing plant tissue. Crosses were made between stably transformed plants expressing functional transposase under the transcriptional control of either the putative AcTPase promoter or the promoter and ®rst intron from the maize ubiquitin (Ubi1) gene, and plants containing Ds-Ubi-bar. In F 1 plants from these crosses, low somatic and germinal transposition frequencies were observed; however, in F 2 progeny derived from individual selfed F 1 plants, up to 47% of the plants showed evidence of Ds transposition. Further analyses of F 3 plants showed that approximately 75% of the transposed Ds elements reinserted into linked locations and 25% into unlinked locations. Transposed Ds elements in plants lacking the AcTPase transposase gene could be reactivated by reintroducing the transposase gene through classical genetic crossing, making this system functional for targeted gene tagging and studies of gene function. During the analysis of F 3 plants we observed two mutant phenotypes in which the transposed Ds elements co-segregate with the new phenotype, suggesting the additional utility of such a system for tagging genes.

No abstract

Instability of transgene expression in plants is often associated with complex multicopy patterns of transgene integration at the same locus, as well as position effects due to random integration. Based on maize transposable elements Activator(Ac) and Dissociation(Ds), we developed a method to generate large numbers of transgenic barley (Hordeum vulgare var Golden Promise) plants, each carrying a single transgene copy at different locations. Plants expressing Ac transposase (AcTPase) were crossed with plants containing one or more copies of bar, a selectable herbicide (Basta) resistance gene, located between inverted-repeatDs ends (Ds-bar). F1 plants were self-pollinated and the F2 generation was analyzed to identify plants segregating for transposed Ds-barelements. Of Ds-bar transpositions, 25% were in unlinked sites that segregated from vector sequences, otherDs-bar copies, and the AcTPase gene, resulting in numerous single-copy Ds-bar plants carrying the transgene at different locations. Transgene expression in F2 plants with transposed Ds-bar was 100% stable, whereas only 23% of F2 plants carryingDs-bar at the original site expressed the transgene product stably. In F3 and F4 populations, transgene expression in 81.5% of plants from progeny of F2plants with single-copy, transposed Ds-bar remained completely stable. Analysis of the integration site in single-copy plants showed that transposed Ds-bar inserted into single- or low-copy regions of the genome, whereas silencedDs-bar elements at their original location were inserted into redundant or highly repetitive genomic regions. Methylation of the non-transposed transgene and its promoter, as well as a higher condensation of the chromatin around the original integration site, was associated with plants exhibiting transgene silencing.

SummaryA system for targeted gene tagging and local saturation mutagenesis based on maize transposable elements (Ac/Ds) was developed in barley (Hordeum vulgare L.). We generated large numbers of transgenic barley lines carrying a single copy of the non-autonomous maize Ds element at defined positions in the genome. Independent Ds lines were either generated by activating Ds elements in existing single-copy lines after crossing with AcTPase-expressing plants or by Agrobacterium-mediated transformation. Genomic DNA flanking Ds and T-DNA insertion sites from over 200 independent lines was isolated and sequenced, and was used for a sequence based mapping strategy in a barley reference population. More than 100 independent Ds insertion sites were mapped and can be used as launch pads for future targeted tagging of genes in the vicinity of the insertion sites. Sequence analysis of Ds and T-DNA flanking regions revealed a sevenfold preference of both mutagens for insertion into non-redundant, gene-containing regions of the barley genome. However, whilst transposed Ds elements preferentially inserted adjacent to regions with a high number of predicted and experimentally validated matrix attachment regions (nuclear MARs), this was not the case for T-DNA integration sites. These findings and an observed high transposition frequency from mapped launch pads demonstrate the future potential of gene tagging for functional genomics and gene discovery in barley.

Summary Efficient negative selection systems are increasingly needed for numerous applications in plant biology. In recent years various counter‐selectable genes have been tested in six dicotyledonous species, whereas there are no data available for the use of negative selection markers in monocotyledonous species. In this study, we compared the applicability and reliability of two different conditional negative selection systems in transgenic barley. The bacterial codA gene encoding cytosine deaminase, which converts the non‐toxic 5‐fluorocytosine (5‐FC) into the toxic 5‐fluorouracil (5‐FU), was used for in vitro selection of germinating seedlings. Development of codA‐expressing seedlings was strongly inhibited by germinating the seeds in the presence of 5‐FC. For selecting plants in the greenhouse, a bacterial cytochrome P450 mono‐oxygenase gene, the product of which catalyses the dealkylation of a sulfonylurea compound, R7402, into its cytotoxic metabolite, was used. T1 plants expressing the selectable marker gene showed striking morphological differences from the non‐transgenic plants. In experiments with both negative selectable markers, the presence or absence of the transgene, as predicted from the physiological appearance of the plants under selection, was confirmed by PCR analysis. We demonstrate that both marker genes provide tight negative selection; however, the use of the P450 gene is more amenable to large‐scale screening under greenhouse or field conditions.