Analysis of transgenic petunias generated by Agrobacterium tumefaciens

Abstract

The aim of this thesis was to study the behaviour of genes introduced into a large number of transgenic petunia plants via Agrobacterium-mediated transformation. A series of initial experiments were performed to establish techniques and media for the in vitro culture of petunia from explants such as protoplasts and leaf segments. These techniques were then used in cocultivation experiments to generate gall tissue and transformed plants using a variety of Agrobacterium strains. During these experiments Horsch et al (1985) published a transformation technique using the cocultivation of leaf discs. This system proved to be effective in producing large numbers of transformed plants. A characteristic of the leaf disc cocultivation technique was the production of a large number of non-rooting shoots, or 'escapes'. Analysis of the cause of these escapes showed that they were not the result of reduced kanamycin selection caused by either neighbouring transformed shoots or the position of origin on the disc perimeter. One hundred and four kanamycin resistant petunia "Mitchell" plants were regenerated in two leaf disc cocultivation experiments using the Agrobacterium tumefaciens strain LBA4404 containing a binary vector pCGN200. pCGN200 carries two kanamycin-resistance genes under the control of the ocs and the 35S promoters. The regenerated plants were analysed for the expression of the kanamycin-resistance phenotype by culturing leaf discs on medium containing kanamycin. Levels of resistance to the presence of kanamycin in this test ranged from completely sensitive to fully resistant. Inheritance of the kanamycin-resistance phenotype was analysed by germinating seeds on medium containing kanamycin. The seeds were isolated from three crosses: a self cross and two backcrosses with a wild-type petunia using the transformed plant as the male and female parent. Approximately half of the plants showed normal Mendelian inheritance patterns for one or two independent dominant genes. In one plant the two genes were inserted at closely linked sites on homologous chromosomes and produced no kanamycin-sensitive progeny in 1894 seeds tested from backcrosses with a wild-type petunia. Six plants had inheritance patterns suggesting that the T-DNA had inserted an essential region of the plant DNA. Forty five of the plants showed lower than expected transmission of the resistance phenotype. Ten plants produced segregation ratios that were not readily interpreted by Mendelian inheritance. There was a reasonable correlation between the level of kanamycin resistance and the pattern of inheritance of the kanamycin-resistance genes in these plants. Southern hybridisation analysis was performed on 96 of the transformed petunias (described above) to study the organisation of the integrated T-DNA. Just over half of the plants contained intact copies of the T-DNA. The most common rearrangements of the T-DNA were simple deleted derivatives that had lost one or both ends of the T-DNA fragment. Left and right border deletions occured at similar frequencies. Three plants contained grossly rearranged T-DNAs of which all expressed the kanamycin-resistance phenotype. However only one of these plants transmitted the kanamycin-resistance phenotype to its progeny. Two plants contained no detectable T-DNA sequences and neither expressed kanamycin resistance in their leaves or progeny. The variation in expression and inheritance observed in the transgenic plants was not caused by alterations in the arrangement of the T-DNA sequences. Rather the variation probably derives from epigenetic effects. The only correlation observed between copy-number of the T-DNA and expression of the kanamycin-resistance phenotype was that most of the plants containing multiple copies of the T-DNA produced anomalous segregation ratios. Some evidence suggests that the kanamycin-resistance gene driven by the ocs promoter is not expressed in roots as well as the same gene driven by the 35S promoter.