Virus-Induced Gene Silencing (VIGS) is a powerful tool for testing gene function in plants with the advantages of avoiding the difficulties of transformation and regeneration. However, relatively few taxa have been studied using this approach, and most that have are in the Solanaceae. Using silencing of phytoene desaturase (PDS), a key enzyme in the carotenoid biosynthesis pathway, as a visual reporter, I tested the feasibility of using VIGS in a wide range of ornamental taxa. A set of degenerate primers was designed to amplify a region of the PDS gene that is highly conserved among species. Amplified PCR product was used to test the effectiveness of VIGS in a range of ornamental species. I observed the characteristic photobleaching phenotype of PDS silencing resulting from the inhibition of biosynthesis of protective carotene on green tissues (leaves or stems) in numbers of the Asteraceae, Leguminosae, Balsaminaceae, Solanaceae and Nyctaginaceae, demonstrating the value of this strategy in a wide range of taxa. In some species, however, the PDS phenotype was not obtained, despite multiple attempts. One example is Mirabilis jalapa (Four O'Clock), an excellent model plant to study flower senescence and abscission. To test the hypothesis that the Mirabilis Antiviral Protein (MAP) may inhibit VIGS-mediated gene silencing in this species, the PDS and MAP genes were co-silenced. The resulting strong photobleaching phenotype confirmed this hypothesis, and suggests a potential strategy for successful VIGS studies in this and other intransigent species.

In the production of potted ornamentals, regulation of plant growth has relied heavily on applying chemical growth retardants, mostly GA inhibitors, to achieve a desirable architecture. Given the high cost, potential environmental impact and possible phytotoxicity of these chemicals, I investigated the potential of using a transgenic strategy for controlling plant architecture, specifically by interfering with GA signaling. The current model of GA signaling suggests that GA binds to a soluble GID1 receptor, which then interacts with a DELLA protein, a repressor of GA responses, releasing repression via targeted degradation of the DELLA protein through 26S proteasome pathway. I hypothesized that I could control ornamental plant height either by silencing the GA receptor (GID1), or by stabilizing the DELLA repressor by over-expressing a defective DELLA protein. Three putative GID1 receptor genes ( PhGID1A, PhGID1B, and PhGID1C) were isolated from petunia. Virus-induced gene silencing of these genes resulted in stunted growth, dark-green leaves and late-flowering. The gai mutant gene (gai-1) encoding a defected DELLA protein was isolated from Arabidopsis. This mutation is due to a 17 amino acid deletion in the conserved DELLA domain of the GAI protein, which results in protein stabilization and extreme dwarfing in the mutant Arabidopsis. The isolated gene was used to produce transgenic petunia plants in which the gai mutant protein is over-expressed under the control of a dexamethasone-inducible promoter. This inducible system permits induction of the dominant Arabidopsis gai mutant gene at a desired stage of plant development in petunia plants by the application of a chemical inducer, dexamethasone (Dex). The induction of gai in Dex-treated T1 petunia seedlings caused significant growth retardation, confirming the hypothesis that a transgenic approach could be an effective strategy for preventing 'stretching' in ornamental plants.